Previously the authors proposed a four stages classification of the fatigue crack propagation, namely the initial, the tensile mode, the shear mode and the final propagation stage. In this report, a method for calculating the crack propagation is proposed based on the authors' classification. A new model is introduced considering the effective stress intensity factor. The initial crack propagation is affected by the compressive stress on the fatigue crack inner surfaces. The propagation in the tensile mode and the shear model stages is mainly controlled by the effective stress intensity factor ΔK_e=K_max-αK_min, and the value of α is expressed by stress ratio, and is independent of the fatigue crack propagation rate and the applied stress amplitudes. The results of this calculation agree well with the experimental results. The usefulness of this calculation in the initial stage is qualitatively shown for the study of non-propagating crack, crack propagation influenced by residual stress and the crack propagation behavior by programed load.

View full abstract

The effects of the stress amplitudes below the fatigue limit on the fatigue lives were investigated by the two steps multi-fold program fatigue tests in which the higher stress was above the fatigue limit and the lower stress was below the fatigue limit. Cylindrical specimens having V-shape groove of the structural carbon steel JIS S35C were used, and the stress concentration factors α of the specimens were selected as 2, 4 and 6. The test results revealed that the cumulative cycle ratios of the program fatigue tests were from 0.2 to 0.6 when α was 2, from 0.8 to 1.2 when α was 4, and from 0.9 to 1.4 when α was 6. The following behavior was also clarified : the fatigue crack was propagated even by the stress amplitude below the fatigue limit and in most cases the crack propagation was stopped as the number of cycles increased. Some discussions were made on the experimental results.

View full abstract

In order to examine the behavior of fatigue cracks in pure iron subjected to rotary bending and cyclic torsion, the copper electroplating method of stress analysis and slip phenomena in the material was applied. The charge in metal properties caused by fatigue cracks had no effect on the propagating stress of cracks. The difference of the fatigue-notch factor between bending and torsion was due to that of both the strain-concentration factor and the strain gradient at crack tips. It was found that in torsion the fatigue cracks propagated when the cyclic shear stress over one grain reached cracks depended on the tensile stress exerted over more than one grain size.

View full abstract

The effects of stress ratio, environment, frequency and temperature on the rate of fatigue crack growth are investigated by use of polycarbonate, polymethyl methacrylate, polyvinyl chloride and ABS resin. It is found that the crack growth rate can be expressed by a power law of the range of stress intensity factors and the constants of the law are strongly influenced by the factors mentioned above. And an empirical relation taking those factors into account is proposed.

View full abstract

Stability analysis of a feed drive servo system of numerically controlled machine tools is important for improving its performance but very difficult since an actual system has a lot of nonlinearities. In this paper, we propose two methods. One is a simple identification method of the nonlinearities and the dynamics of hydraulic closed-loop digital servo system used in an actual NC machine tool, which utilizes a measured hydraulic motor input pressure. The other is a semi-graphic solution method using a describing function on the stability problem of a system which has more than two nonlinearities such as DA conversion nonlinearity, nonlinear friction, and nonlinearities between spool displacement in servo valve hydraulic motor input pressure and flow rate. Using these methods, it has been shown, the nature of stability and the mechanism of self-excited oscillations observed in an actual system can be clearly explained.

View full abstract

Some experimental investigations on unstable performances in partial flow rate regions of a single stage axial-flow compressor and its rotating stall are described. In order to explain the unstable operations related with rotating stall, the influences of hub-tip ratio and rotor-stator distance on the performance curves and rotating stall characteristics are discussed. Performance curves, total pressure distributions under different operating conditions, size of stall cells and stall velocity are shown, Many types of rotating stalls, such as large stall, small stall, partial span stall and full span stall are investigated with different hub-tip ratios and stator positions. When large stall occurs, the discharge pressure discontinuously decreases and a slope of the performance curve becomes positive. In the case of small stall, however, the pressure decreases slightly and continuously. The partial flow rate performance is severly affected by the size of stall cell.

View full abstract

Two experimental investigations of the off-design operations of axial-flow compressor are described. One of them is an investigation of influences of impeller load on both partial flow rate performances and the rotating stall characteristics, the other being an attempt to make clear the mechanism of the unstable performance related with rotating stall. The width of the rotating stall regions and a decrease in discharge pressure in the partial flow rate region become small as the impeller load decreases. When the impeller load becomes so small that rotating stall completely disappears, the slope of the performance curve becomes negative at all flow rates. It is clear that the reverse flow of rotating stall (large stall) reverses from delivery to suction side across the rotor. The mechanism of the reverse flow is discussed and it is shown that the reverse flow is the cause of the unstable performance.

View full abstract

Experiments on friction of parallel coaxial discs rotating in a cylindrical vessel with some surface roughness were carried out under various radial, axial clearances and disc intervals. The main results obtained are summarized as follows : (1) Resisting coefficient of a special single disc estimated as if each surface of the disc were facing the vessel surface fixed and rotated with same speed respectively, is larger than the value of an enclosed single disc. (2) Effects of radial clearance for resisting coefficient decrease with an increasing disc interval, and disappear when the ratio of disc interval to disc dia. is 0.6∼0.8.

View full abstract

The slow motion of a viscous incompressible fluid in the space bounded by two long parallel eccentric circular cylinders is studied. The outer one is stationary while the inner one is rotating steadily. Approximate solutions of the Navier-Stokes equations are obtained for the two dimensional flow. If the centers of cylinders approach each other, these solutions may coincide with a stream function between concentric rotating cylinders. Also, if the gap between the cylinders is small, they should coincide with the Reynolds approximate solutions. Authors' theoretical solutions are confirmed experimentally when rotating Reynolds number is less than 7.

View full abstract

This paper treats the performance of a low head airlift pump used to lift solid particles. The flow inside the vertical pipe of the air-lift pump is a three-phase flow, i.e. air, water and solid mixed flow. It can be analysed based on the existing theory for an air-water two-phase flow. By coupling the momentum equation of flow and the equation of motion of a single solid particle, the problem can be solved numerically. The results of computation agree with those of experiments quantitatively. The performance of a typical air-lift pump for solid particles is calculated using the present method.

View full abstract

In this report is studied a fluidic oscillator, made of a nonmemory bistable amplifier, with loads and feedback tubes. A mathematical model for describing approximately the performance of the oscillator is presented. Analytical results derived from the model are examined by experiments. The results obtained are summarized as follows : (1) The model explains the performance of the oscillator fairly well. And the relation between performance and geometry of the oscillator is clarified. (2) Analytical results concerning the condition for sustained oscillation and the frequency of the oscillation show good agreement with experimental data.

View full abstract

The system of cylinder-rows is favored in heat-exchangers or radiators, but estimation of its radiative transfer is often unavailable due to insufficiency of data provided in few reports. The purpose of this study is to identity the characteristics of radiative transfer in such cylinder-row systems. Analysis is made on two-row gray circular cylinders arranged on square-centers to obtain local radiative fluxes. Then the influences on radiative transfer are found of emissivity and the center-to-center distance of cylinders. Approximate solutions are additionally evaluated to compare with the numerical solutions. The same procedures are employed in one-row cylinders to see the influence of the number of rows. In black, semi-infinite-row cylinders (arranged infinitely from the first row on one side), variations in radiative heat rate are evaluated in several rows from the initial, in the range where the heat transfer is significant.

View full abstract

In this report, apparent thermal conductivity perpendicular to the fiber of compound material, which contains parallel fiber elements of irregular distribution, is discussed. (1) Concerning many compound materials which are taken at random using random numbers, the values λ/λ (the ratios of apparent thermal conductivities λ and conductivities P of models of parallel arrangement λ_P) are shown to be approximated by a linear function of η (the volume ratio of discontinuous phase), for η = 0∼0.2. (2) Defining ξ=λ_P/λ_h, where λ_h is the value for uniform fiber distribution, the value of ξ is discussed statistically. (3) Considering (1) and (2), the approximate equation of apparent thermal conductivity is derived as, λ/λ_c = [(1+Fη)ξ]/[(λ_c/λ_d)η+1-η], F=1.08-6.8/[(λ_d/λ_c)+5.3], (η=00.2, λ_d≥λ_c, l/d≥10) which is shown to agree within an error of 8% with the experimental data of electrolytic bath.

View full abstract

As a serial study on the damping capacity of a bolted joint, in here effects of the machining method and roughness of joint surfaces, joint materials and also interfacial layers, on the damping have been investigated. By measuring the logarithmic damping decrement and the natural frequency of bolted cantibeams, how the damping and its amplitude dependence, and also the optimum preload, change themselves with different conditions of joint surfaces, has been also discussed. From the results it is known that the contact condition of joint surfaces has large effects on the damping capacity, but a little effect on the natural frequency. In the metal contact condition, the micro-slip on joint surface caused by the plastic deformation or the tearing of seizured points, is of importance to know the behaviour of damping. In the oiled joint, the increase of damping can not be always expected even if high viscosity oil is applied to joint surfaces.

View full abstract

There have already been many studies on the pitting of power transmitting gears. However, many problems seems to be left still unsolved. This report deals with the experimental investigations an the correlation of the data by means of a statistic method, cosidering the following two aims : (1) To collect information on the effect of factors affecting the occurrence of the pitting of power transmitting gears. (2) To obtain the experimental data with statistical confidence on the pitting strength of power transmitting gears. (2) To obtain the experimental data with statistical confidence on the pitting strength of power transmitting gears. As the result of preliminary study, it was made clear that the distribution of pittings follows a logarithmic normal distribution. Moreover, investigation was made on the factors influencing the pitting formations.

View full abstract

In the case of applying a lever which swings about the horizontal axis by propotional control of an electric servomechanism to a manipulator's arm etc., it is difficult to position the manipulator's arm correctly according to the unbalance due to the grasping and releasing an object. In this paper, a balancing device using a movable counterweight is proposed for the purpose of positioning the arm correctly, and its characteristics and stability are studied. As a result of analysis, it is confirmed that such a balancing device is available for the manipulator's arm and automatic balancing device.

View full abstract