There are many problems on behavior of a fatigue crack that are not still revealed by many studies with an electron microscope. Macroscopic behavior of the fatigue crack cannot be completely explained by its electron microscopic one. Medium scopic observation between the macroscopic and the electron microscopic one should be made to fill the gap between them. In this investigation, the vicinity of the fatigue crack tip is observed every several cycles of stress by an optical microscope of 280 magnification and the fracture surface is observed by the electron microscope. The relation between an optical microscopic observation and an electron microscopic one is discusssed and the following conclusions are obtained : (1) It is observed that a fatigue crack intermittently propagates. So, there are dormant terms in a fatigue crack propagation process.(2) The dormant terms ΔN and a crack length Δl which intermittently propagates after a dormant term depend on material and applied stress. The former is determined by alternating stress and the latter determined by maximum stress.(3) Striations are found on the fracture surface on which a fatigue crack intermittently propagates.

View full abstract

The rotating bending and axial fatigue tests of plain and notched 7 : 3 brass specimens were carried out.The main results obtained are as follows.(1) The characteristic in fatigue of notched specimens under rotating bending stress is similar to the one under completely reversed axial stress, and the root radii of notches at the branch point ρ₀ (branch point means the critical point for a non-propagating crack to develop) are about 0.6 mm for both cases.(2) The relations of σ_w1/σ_w0 and σ_w2/σ_w0 (σ_w1 : limit stress for crack initiation, σ_w2 : limit stress for complete fracture in the range of non-propagating crack existing, σ_w0 : σ_w1 of plain specimens) to stress concentraiton factors α are almost the same in rotating bending fatigue of 7 : 3 brass and 0.13 % C steel.(3) In axial fatigue, the maximum stress repeated at a notch root at the crack initiation limit ασ_w1 becomes almost constant irrespective of mean stress, if the stress gradient χ is constant.(4) In axial fatigue, ρ₀ increases as mean stress decreases. This means that σ_w2 increases as mean stress decreases.

View full abstract

In the present paper, some basic propertiesd of generalized solutions for the equation of 3-dimensional elastostatics were considered and the following results were obtained : (a) The elementary solution of elastostatics (Somigliana's tensor) satisfies the basic equation in the sense of distribution. (b) The basic formulas for the theory of elastic pontentials are derived from the convolution of elementary solution and a distribution. (c) If body force is absent, any generalized solution for the bounded domain is infinitely differentiable. (d) The displacement solution for the problem of uniform extension of an infinite body is explained as a generalized solution in the space of slowly increasing distribution. (e) If a displacement solution of the equation with no body force is bounded in the whole domain R³, then it is a constant.

View full abstract

The influence of polygonized sub-structure on flow stress of pure aluminum (99.9%) was investigated.As flow stress increases parabolically with the inverse square root of sub-grain size, Petch's relation does not hold true between flow stress and sub-grain size. It is clarified that not only sub-grain size but also misorientation across the sub-boundaries has great influence on flow stress.In addition, it is observed during the experiment that flow stress σ_f increases with the square root of excess dislocation density D_b of the sub-boundary, yielding the relation σ_f=σ₀+α'GbD_b^1/2.The effect of sub-structure on flow stress appears to be better explainable in terms of the work-hardening theory which is inclusive of the interation of dislocations rather than Petch's pile-up theory.

View full abstract

The basic equations are proposed for plastic deformation of porous metals such as sintered metal powders or metal powder compacts. The porous metal is assumed to be made up of a homogeneous metallic material and unformly distributed pores. The following equations are obtained from the Levy-Mises equatioins, (dε₁-dε₂)/(σ₁-σ₂)=(dε₂-dε₃)/(σ₂-σ₃)=(dε₃-dε₁)/(σ₃-σ₁)=(3fdε_υ)/(2σ_m/f)=(3dε_eq)/(2γσ_eq) where γ is the relative density of the porous metal, f a function of γ, the subscripts 1, 2 and 3 represent the principal directions, σ_m=(σ₁+σ₂+σ₃)/3 and dε_υ=dε₁+dε₂+dε₃. The theory shows a good agreement with the experimental results for sintered copper powder compacts.

View full abstract

The flat plane speaker has wide sound radiation area making its tone characteristics preferable to the conventional type speakers. However, the main defect is its multiresonance of the vibrating membrane. After the analysis of flat membrane vibrations, a new driving method is proposed for eliminating the multiresonances, that is, if the membrane were driven by the distributed force proportionate to the fundamental mode amplitude, no multiresonance would take place. Mean while within some restricted frequency range, the uniresonance is expected from some discrete concentrated driving points. In the preceding paper, this driving method has been applied to the cases having easily analizable boundary conditions. On the contrary, this paper includes the application to three complicated problems, entirely clamped, free and mode degenerated. Comparing them, the entirely free boundary condition is found theoretically best, however, the problem is how to hold the free membrane in the space. As the result of various experiments, the boundary condition almost the same as the free boundary but having a small viscous damping element which makes it possible to solve the suspension problem is found.

View full abstract

In the previous paper the isolation of random vibration was considered when an ideally white noise over the whle frequency range was loaded on the movable foundation of a two-degrees-of-freedom linear system.In many cases, however, frequencies are practically limited in a specially-fixed range. Then in this paper the case of narrow-band process is dealt with. First the case of white noise is investigated from the rms value of absolute accelerations and relative displacements and secondly the case of ramp noise, whose example is a power spectrum in the roughness of road, is considered and we shall compare the case of ramp noise with that of white noise qualitatively and quantitatively, where the introduction of equivalent spectral density is tried in order to correlate both cases.

View full abstract

This paper describes theoretically the development and separation characteristics of a laminar boundary-layer flow over a flat plate placed in a uniform shear flow which is accompanied with a streamwise pressure gradient. The main flow is assumed as u'=U'(x')+wy' with U'(x')=U₀{1-(x'/l)ⁿ}, where w, U₀ and n are constants. The govering equations have different forms in the leading edge region and in the vicinity of the separation point; the former is solved by the perturbation method, and the latter by the finite-difference method. The result shows that the separation of the boundary layer is affected considerably by the existence of w. It delays when w is positive, while it proceeds upstream when w is negative. The point of separatioin can be expressed as an expotential function of α=KR^-1/2 for small values of α, where K=wl/U₀, R=U₀l/v and v kinematic viscosity. The analysis also shows that the effect of the external vorticity on the separation becomes less significant as the pressure gradient becomes smaller.

View full abstract

An apsect of laminar wake of a symmetrical two-dimensional body placed in a uniform shear flow is described theoretically within the framework of the second-order boundary layer theory. The effect of the external vorticity involved in the oncoming flow is treated as a small perturbation to the case of uniform flow. An analytical solution for the far wake is obtained together with an approximate numerical solution for the near wake behind a flat plate aligned parallel to the oncoming flow, showing the overall development of the wake in the uniform shear flow.The analysis shows that the spreading of the wake is greater towards the side of smaller velocity than the side of larger one. However, the width of the wake defined as the distance between two points at which the streamwise velocity defect amounts to one-half of the maximum defect is found to be little affected by the presence of the external vorticity.In the present flow configuration, the displacement effect also appears as another second-order term which is found to be less significant in the far wake region than the effect of the external vorticity.

View full abstract

In order to extend the measuring range of the float-area-type flow meter, a float with a hole was tested. The flow entering the tapered tube of the conventional float-area-type flow meter is passsed, in this case, both the clearance between the tapered tube and the float and the hole of the float, and the total flow rate is indicated by the position of the float. The flow rate characteristics of this type of flow meter was theoretically and experimentally investigated.The curve for the resulting flow rate characteristics was found to shift almost parallel to itself with an increase in the diameter of the hole. Typical basic data for designing this type of flow meter were presented.

View full abstract

Concerning the characteristics of frequency and amplitude of pressure surge caused by whirl flow from the runner outlet of a Francis turbine, experimental investigations by model tests were made with the folloowing results.The relation between frequency under general situation and remarkable pressure surge and specific speed with wide range was made clear using the circulating velocity of flow from runner exit.A practical interpretation about tendency of increase and decrease of amplitude according to operating conditions and amplitude characteristics affected by cavitation coefficient was obtained.

View full abstract

The characteristic equatioins in which the inertia and compressibility of oil and the squeeze film effect are considered, are derived for a hydrostatic thrust bearing supporting a constant load and intermittently being supplied with pressurized oil as square wave form.Meanwhile, these characteristics are examined by experients, to confirm the relation between the mean value and its variations of the oil film thickness and the pressure wave frequencies.Conclusions of this study can be summarized as follows;(1) The mean value of the oil film thickness is functionally determined by the value and ratio of the supply pressure and the restriction coefficient of the choke, if the supply pressure is always higher than the balancing pressure.(2) The squeeze film effect evidently appears in low frequency region, but in high frequency region and with strong restriction effect it diminishes because of small variation of the oil film thickness.

View full abstract

In order to solve two-dimensional transient heat conduction problems by finite difference method with required accuracy, authors discuss to divide the temperature field using three kinds of meshes along the boundary, that is, (1) rectangular meshes, (2) meshes divided by cicles and radial lines, and (3) triangular meshes. In this report, the following discussions are made, concerning rectangular meshes.(1) Explicit-type difference equations approximating seven boundary conditions which are usually used in heat conduction theory, and those stability conditioins are revealed.(2) A difference equation, which converges more rapidly than the ordinary difference equation, is proposed.(3) A Flow-Chart, by which the numerical solution of required accuracy is obtained by controlling the parameters of at<min>/(ΔL)² and t<min>/Δt, is shown. And its effects are illustrated by an example.

View full abstract

Burn-out phenomena are studied for two special systems of boiling which are both realized in the saturated nucleate boiling of water at atmospheric pressure on a horizontal 10mm dia. copper heated surface.First, it is attempted by mechanical means to cause a forced collapse of the vapor masses which are generated periodically on the heated surface. In this case, burn-out heat flux q_c increases, but the velocity of fluid movement concerned with the collapse of vapor mass is restricted so that the increase of q_c is limited.Second, the heated surface is supplied with liquid by means of a small liquid jet. In this case, burn-out appears with a mechanism which is different from ordinary pool boiling. Splashing of liquid due to the violent effusion of vapor from the heated surface interrupts the part of supplied liquid in reaching the heated surface so that the quantity of liquid on the heated surface balances finally the heat flux, and burn-out occurs. For the burn-out of this type, q_c can be increased steadily by increasing the velocity of liquid jet.

View full abstract

The process of drop growth due to both direct condensation of vapor onto the drop and coalescence with neighboring droplets during dropwise condensation is investigated.Obviously it is the process of coalescence that largely controls the rate of drop growth and, eventually, the rate of condensation, but the random or stochastic nature of this process has prevented us from conquering the problem.In this paper, some analytical considerations are made of this complicated process, and mathematical expressions which relate the rate of growth of each individual droplet, together with the distribution of drop size over the condensing surface, to the macroscopic heat-treansfer coefficient are derived.Also the method of computer simulation, which the authors have developed in their previous works, is successfully employed to determine the parameters of the growth process.

View full abstract

Equations for calculating the flow rate ratio of each branch pipe fitted on a confluent header with a short pitch are derived taking the pressure variation of a two phase mixture in the header into acount. The dimensionless confluent flow factor introduced in those equations has a similar form to one for a single phase flow and characterizes the flow rate ratio of branch pipes.The pressure loss of the branch pipe where liquid is heated uniformly along the pipe length and flows out in a state of a two phase flow is affected by the quality of flow at the outlet of the branch pipe. The flow rate distribution in branch pipes is measured for air-water mixtures by the equipment arranged to give the above pressure loss characteristics in the branch pipe. From the experimental results, the confluent flow factor is evaluated and compared with one for a single phase flow. The flow rate distribution in branch pipes predicted from those equations is in fairly good agreement with the experimental results.

View full abstract

A new linkage device has been developed for the measurement of piston temperatures in high speed engines with a high degree of accuracy and long durability. The piston temperatures have been successfully measured under various driving conditions on ordinary roads by this device. As a result of comparing these temperatures with the laboratory ones we reached the following conclusion : i) During driving, piston tmeperatures fluctuate, responding rapidly to changes in engine conditions. The temperatures change frequently and large over a range of 100°C, particularly on highways.ii) During acceleration, the longer the "knock" persists, the more the temperature rises. In this case, the highest temperature recorded is about 340°C.iii) Generally, the temperatures obtained nearly steady state conditions on the road agree well with those obtained in the laboratory.

View full abstract

Many mechanisms for imparting intermittent motion have already been devised. Among these mechanisms, the Geneva mechanism is most commonly used, because of its simplicity. However, it is inferior to a barrel cam mechanism from the view-point of dynamic characteristic.Therefore, to remove this fault, the author has tried to rotate the driving shaft of the Geneva mechanism nonuniformly, and prposed a new intermittent mechanism which is named "compound intermittent mechanism".In this paper, the characteristic of the compound intermittent mechanism has been discussed theoretically. The paper is summarized as follows. The compound intermittent mechanism is superior to the Geneva-gear train, which is commonly used, in the next points. (1) (Dwell period./index period) can be easily changed. (2) The indexing angle is precise. (3) The fluctuating input torque is small.

View full abstract

It is necessary to have a knowledge of the pressure distribution under dynamical conditions for the analysis of dynamical characteristics of a rotor bearing system.For this purpose, in this paper, the dydnamical pressure distribution in the aerodynamic journal bearing is analyzed using time-dependent Reynold's equation for small circular motion of journal center.To examine the reliability of the present solutions, the angle was calculated between the direction of miniumum thickness of the gas film and that of the maximum dynamical pressure under the condition of synchronous whirl.By comparison of the calculated values with the experimental ones, it was made clear that, when bearing number becomes large, it is necessary to consider the time-dependent effect on the dynamical pressure distribution.

View full abstract

The loss energy of a jointed cantibeam may be derived from a microscopic slip and a normal pressure on the interfaces. It is however very difficult to calculate theoretically the energy dissipation between the joint surfaces, because a characteristic of the micro-slip on the interfaces has not been fully clarified yet.In this study the mechanism of the energy dissipation and also the ways of theoretically calculating the damping ratio of the jointed cantibeam have been investigated by considering experimental results of the characteristic of static micro-slip between the jointed surfaces.From the results it is clarified that the damping ratio of the jointed cantibeam can be well calculated by using the equations proposed in this paper, and how the characteristic of micro-slip influences the behaviours of the damping ratio.

View full abstract

The noncircular gears have been used in the flowmeters, pumps, etc., but they were used, for the mechanisms with nonuniform motion, to make use of the monotonous angular relations between the input and output axles by fitting them to the ordinary gear trains.In this paper, the planetary gear mechanisms were composed with two pair of gears, one pair with noncircular gears and the other with circular gears, and these mechanisms were analyzed as a function generator. Then, the mechanisms were classified into swing, swing and rotate, and intermittent and rotate motion, and the restrict conditions were obtained, and examples of synthesis with the displacement curves of practical use were shown. Furthermore, the mechanisms were produced by way of experiment, and by the experiment, it was shown that the noncircular planetary gear mechanisms with nonuniform motion make a function generator with good characteristic.

View full abstract

A number of surface-hardened gears are usually used for heavy load high speed gears in which scoring is a serious problem. When the scoring factor on these surface-hardened gears are to be determined, it is necessary to examine both the variation of the thermophysical properties in the surface layer due to surface-hardening and the effect of the variation of the thermophysical properties on the equation of the temperature rise of the surface.We have obtained the following results through both the calculation of the surface temperature rise caused by frictional heat of a surface-hardened gear and some measurements on the thermal conductivity in the surface layers of surface-hardened gear teeth.(1) The thickness of the surface layer of the surface-hardened gear, such as a high-frequency induction hardened gear and a carburized hardened gear, is usually greater than the width of Hertzian contact zone. Therefore, when the surface temperature rise of the surface-hardened gear tooth is calculated, the thermal conductivity in the surface layer of the surface-hardened gear is used for calculation.(2) After heat treatment of hardening or carburizing, the thermal conductivity of the material, such as S 45 C steel or SCM 22 H steel, falls remarkably.

View full abstract

A kinetic equation of machine tool slideway motion is obtained considering slideway conditions, and the coefficients of the equation are expressed as a state function which is related to driving velocity and film thickness in a steady state. The transient response of the state function in such a nonsteady state as stick-slip motion can be approximated by the time lag of first order.The analysis of the kinetic equation by local linearization supports the following experimental phenomena. Namely, stable sliding occurs even in the negative slope range of characteristic curve of friction, and the larger the rigidity or oil viscosity, the wider the stable range is. Since there are many factors which influence the coefficients of the equation, it is difficult to solve the equation analytically. The simulation on an analogue computer, therefore, has been attempted. The results thereof show that the obtained equation represents the actual behavior of sliding motion well.

View full abstract