Elastic deformation of polycrystals under uniaxial and shear loadings is studied with the application of the idea of constraint ratio which represents the mode of deformation of polycrystals. Local deformation behavior of polycrystals is analysed using a plane rectangular model of polycrystals and the finite element method, where elastic constant of each grain is assumed to be represented by Young's modulus and Poisson's ratio. Discussions are made on the relation between the calculated local behavior and the overall deformation behavior of polycrystals. The value of constraint ratio of a grain in tension changes from constant stress to constant strain according to the changes of elastic constants of the neighbor grians. The change of constraint ratio is also affected by the aspect ratio of grains. The mean equivalent constraint ratio for polycrystals is defined when each grain in polycrystals has a different value of the mean constraint ratio. The effect of Poisson's ratio on deformation behavior of grains is found to be relatively small in general, while the effect of the transverse Young's modulus perpendicular to the stress axis is comparatively large. The mean constraint ratio in shear deformation takes its minimum value for a certain aspect ratio of grains. Discussions are dame on the relation between analytical and experimental values of elastic constants of polycrystals and those calculated from elastic constants of single crystal.

View full abstract

This paper contains a three-dimensional solution for the stresses in an elastic thick plate containing an eccentrically located spherical cavity under uniaxial tension. The solution is represented by a combination of the solution which is regular in the outside of a spherical cavity and the solution which is regular in an infinite thick plate, and is deduced with the aid of the six stress functions used in the author's previous papers. Numerical results are given for the eccentric distances c=0.25, 0.5 with the ratios of cavity diameter to plate thickness varying 0.05∼0.45 and 0.05∼0.35 respectively, and the stress distributions around a spherical cavity are shown graphically. Comparison is made between the axisymmetric solutions obtained by C. B. Ling et al. and ours.

View full abstract

This paper contains a three-dimensional solution for the stresses and deformations arising in an elastic thick plate with a symmetrically located spherical cavity under transverse bending. The solution is represented by a combination of the solution which is regular in the outside of a spherical cavity and the solution which is regular in the infinite thick plate, and is deduced with the aid of the six stress functions used in the authors, previous papers. Numerical results are given when the ratios of cavity diameter to plate thickness are 0.2, 0.4, 0.6, 0.7 and the stress distributions around a spherical cavity are shown graphically. Comparison is made between the axisymmetric solutions obtained by C. B. Ling et al. and ours.

View full abstract

The authors analyze the axisymmetric stress distribution in an infinite elastic solid containing a flat annular crack under torsion on the basis of three dimensional theory of elasticity. Assuming that the circumferential displacement on the crack surface is continuous, we may represent it by Fourier series and reduce the problem to the solution of an infinite system of simultaneous equations. The radial distributions of the displacements and the stress components and the variations of the stress intensity factors at both crack tips with the ratio of the inner radius to the outer one are shown graphically.

View full abstract

By using an electroplated nickel, a fundamental research for measurement of cyclic stresses in high temperature range is carried out. The plating used consists of two layers, namely, a preliminary deposit of alkaline copper in a copper cyanide bath and a dull nickel deposit in a Watts bath. It is suitable for the present purpose that the current density of nickel deposits is taken to be below 0.5 mA/mm². In the high temperature range up to 200°C, grains of deposited nickel grow large when the deposits are subjected to a cyclic shearing strain of more than a certain magnitude. As a result of examination of the relation between the shearing strain with which grains begin to grow and the number of cycles, it is found under the plating conditions tested that the available testing temperature for the method is within the range of 100 ∼230 °C.

View full abstract

A fundamental research for measurement of cyclic stresses in high temperature range is performed by applying a Watts-type nickel electroplating. The following results are presented. For the plating solution and plating conditions used, the range of applicable temperatures is from room temperature to 190°C. Though the desirable number of repetitions N is in the range of (0.5∼4)×10⁶, it is more appropriate for the accuracy of measured values to set N= (2∼3)×10⁶. Adding a small quantity of gelatin to the nickel plating solution, the applicable temperature range is extended toward higher temperature side. As a result of measurement of the stress-concentration factor in a circular shaft with a flat part under torsion, the measured values at each testing temperature are in good agreement with the theoretical result and it is verified that the present technique has sufficient accuracy.

View full abstract

The subsequent yield surfaces after preloading along various strain paths of straight lines with or without a corner are obtained for an aluminium alloy whose stress-strain relation is not determined uniquely by the effective stress-strain value of Mises-type. Tests are made by subjecting thin-walled tubular specimens to combined axial load and torsion. The shapes of the yield surfaces are discussed in the deviatoric stress plane. Yield criterion is derived experimentally as a function of the length and the direction of the prestrain path. The criterion can express a rounded corner in the preloading direction, the flattening at the opposite side of the preloading, and the distortion caused by the difference of flow stress values between axial load and torsion, appearing in the yield surfaces obtained by experiment.

View full abstract

An impact attenuation system whose idealized static load-displacement curve consists of an elastic range and unrestricted plastic flow is represented by a dynamically equivalent model which takes into account elasticity, viscosity, plasticity and strain-rate sensitivity of the structure, and its dynamic response to the impact by a moving mass is analyzed. The theoretical results are in fair by good agreement with the experimental results on mild-steel simply supported beams. An analysis is also made by regarding the equivalent model of the impact attenuation systems as Bingham's model and the results are compared with the theoretical results by the author's model and the experimental results. Moreover, comparisons between the experimental and theoretical results on plastic impact of beams given by Parkes and the theoretical predictions by author's model are made.

View full abstract

This paper discusses the problems of flexural wave propagation in an infinite bar with a circular arc bend, in an infinite L-type bar with a sharp bend and in an infinite Z-type bar with two bends when excited by an incident wave. A new improved theory taking into account the effects of the extension, rotatory inertia and shearing deformation is advocated. The bending moment concentration factors for these types of bars at their important cross sections are calculated and the differences of the dynamical behaviors among these bars are clarified.

View full abstract

Following the first report, steady impact vibrations in a mechanical system subjected to various harmonic exciting forces are analysed by a perfect Fourier series method reinforced by convergent improvement by means of a series transformation. In this report, the hysteresis loop characteristics for force of restitution are assumed to be composed of four straight-line segments. Under this assumption, contrarily to the usual investigations, the coefficient of restitution is not constant and the duration of collision is not zero, but they depend on the colliding velocities and the elastic and plastic behaviors of colliding bodies. The resonance curves, namely curves of the amplitude (X_max-X_min) /2e_o versus the frequency ratio Ω and the curves of the effective phase lag angles δ versus the frequency ratio Ω for several exciting ratios q/ke_o taken to be constant are obtained by our method utilizing a digital computer. The comparison of the results of numerical computations by the theory with those from an analog computer yields good coincidences.

View full abstract

In the optimization of chemical processes, which contain a number of control inputs, Dynamic Programming, Maximum Principle and Direct Search Methods can be used in some particular processes. In this paper, the Direct Search Methods and modified systematically and combined with five new techniques, so that four Mixed Strategies are formulated. With these Mixed Strategies, the optimization problems of some chemical processes are solved, and the effectiveness of these Mixed Strategies is confirmed.

View full abstract

In this report, results of an experimental investigation which was performed to eliminate the streamwise vortices arising in a nominally two-dimensional turbulent boundary layer of which behaviour was examined in detail in the authors previous paper, are presented. Though there may be two procedures of obliteration of streamwise vortices, the present work is concerned with the method to place the vortex generators like roughness element at the leading edge of a flat plate inducing the counter rotating vortices against pre-existing vortices in the boundary layer. Using this set, the authors measured the downstream development and uniformalization of the spatial variations of the boundary layer characteristics. From these results, it was revealed that the streamwise vortices were extremely stable, and large vortex generators must be used to eliminate these vortices. In addition, these streamwise vortices, which disappeared once in the course of the boundary layer development, exhibited a tendency to appear again in the downstream region. And a weak variation of velocity with large wave length at the outside of the layer produced by the protuberances was observed.

View full abstract

The pressure distribution and the boundary-layer flow on the surface of rotating cones or cone-cylinders having vertex angles of 15°and 30° in a uniform stream were measured, and the effects of the vertex angle, the velocity of uniform stream and the presence or lack of a cylindrical body were studied. It was concluded that the pressure on the rotating cone is not changed by the velocity of the uniform stream and the rotation, and the pressure over the rear part of cone decreases when the cylinder body is present. The relation between velocity components u and v in the boundary-layer depends upon the ratio (λ) of the circumferential velocity to the velocity of uniform stream. But the curves expressed by this relationship are similar for the same value of λ and independent of the vertex angle, the presence of lack of cylinder body, and the laminar of turbulent state. The value of λ at the transition point increases as the vertex angle and the negative pressure gradient increase.

View full abstract

Drag reduction phenomena of liquid flows due to an addition of polymer molecules are caused by an interaction of the elongational viscosity of the polymer solution and the elongational-type secondary flow of the base flow. In transition form laminar to turbulent flow on a flat plate boundary layer the elongational viscosity acts on the secondary vortices appearing by amplification of Tollmien-Schlichting waves and extends the stable region of Tollmien-Schlichting waves. At the same time the wave number corresponding to the critical state of neutral stability for the waves becomes larger with an increase in elongational viscosity. By this analysis it is suggested that the addition of polymer can delay the turbulent transition of boundary layer flow.

View full abstract

The effects of surface roughness on the frictional moment of a rotating cone were investigated experimentally, when the cone was rotated in a casing of which conical surface was parallel to that of the rotating cone. Torque measurement was made for six vertex angles θ, from 30° to 180°. Results obtained are as follows. (1) The roughness of the rotating conical surface has a stronger effect on the frictional moment than that of a stationary one, and this trend is marked at large clearance ratio s'/R. (2) When the roughness of the rotating conical surface is equal to that of the stationary one, the moment coefficient C_M is independent of the Reynolds number R_e at sufficiently large R_e, namely within a completely rough regime. (3) In the completely rough regime, the C_M-values for θ<₌ 120° are larger than those for θ>₌ 150° at a relatively small s'/R. This difference of C_M due to θ increases as the relative roughness k/R increases. (4) Within the completely rough regime, C_m is found to be expressed by the formula C_M = {Alog(R/k)+B}^-2 where A and B depend on θ and s'/R.

View full abstract

Pressure and velocity of pulsating turbulent air and water flows in circular pipes generated by rotary valves were measured by the use of a strain-gauge-type pressure transducer and a hot-wire anemometer, respectively (0.0432 <₌ f <₌ 48.0 Hz, 7740 <₌ Re <₌ 95900). Comparisons of experimental data with approximate solutions derived from a four-region model for a pulsating turbulent flow in a circular pipe containing a slightly compressible fluid show that : (i) Cross-sectional distribution of time-averaged velocities agrees well with that of steady flow velocities. (ii) Cross-sectional distribution profiles of oscillating velocity coincide well with the above-mentioned theoretical ones, and therefore, depend only on the Reynolds number and the dimensionless angular frequency w'=R²w/v, where R is pipe radius, w angular frequency, v kinematic viscosity. (iii) Axial distributions of oscillating pressures and velocities on the pipe axis also agree well with the above-mentioned theoretical ones within the frequency range of 0.1 <₌ f <₌48.0 Hz.

View full abstract

An experimental study on the confluence of two two-dimensional incompressible turbulent jets, issuing from nozzles with angle θ₀ against the nozzle exit plane, has been presented. The axial momentum and the energy flux of the jets along the z-axis perpendicular to the exit plane are calculated by using the measured values of velocity and pressure distributions. It has been shown that the axial momentum flux along the z-axis decreases owing to the confluence, that the energy flux also decreases remarkably owing to the confluence, that the energy flux also decreases remarkably owing to the confluence, and that the decreased values are approximately 40% of the efflux energy of the jet, independent of the ratio of a distance between two nozzles to the nozzle width and θ₀. A method of predicting the velocity distribution also has been proposed. The predicted values agreed well with the experimental results.

View full abstract

Experiments have been made in a shock tunnel non-equilibrium MHD power generator using non-seeded argon plasma. A linear generator with constant cross section and a diverging generator with diverging cross section have been used for the experiments. The main purpose of the experiments is to observe the effects of J × B force on the plasma flow when large powers are extracted from the generator. As the result, the MHD power generator performance is shown to be well described by the two temperature theory with considerations of Coulomb collision, the segmentation of the electrodes, the electrode voltage drops and J × B force effects. In the linear generator, the reflected shock wave was observed at the interaction parameter 5 and the distribution of the steady current was affected by the shock wave. The enthalpy extraction efficiency reached 4 % in the linear duct. When this efficiency is above 1%, J × B force effects become appreciable.

View full abstract

The effect of radiation heat transfer in the measurement of thermal conductivity for the semitransparent medium is examined, in general, as the parameters of the radiation conduction parameter, the optical thickness and the surface emissivity. The region where the superposition approximation of radiation and conduction can be applied is clarified, and then, the region where the radiation heat transfer can be neglected within an error of 5% and the region where the optically thick approximation can be applied within an error of 5% are clarified. Lastly, these examinations are applied to soda-lime glass.

View full abstract

A new analytical method is presented for the synthesis of a plane four-bar mechanism the segment on whose coupler link may pass through five precision positions specified in a stationary co-ordinate system. The numerical procedure is established in such a concise form that consists in solving three systems of five linear equations and a quaritic equatuin. This method is extended to the synthesis of such a mechanism that the higher order differential coefficients of three position variables of the segment on the coupler link at precision positions may coincide with the specified values. Moreover, combining some calculations on the basis of the Roberts-Chebyshev theorem with the numerical procedure, this analytical method may be used as a general precision derivative approximation for designing a path generating mechanism or a function generator.

View full abstract