Composite models can be made of epoxy resin and metal stringer by Araldite adhesive bonding with few initial bonding stresses, and having semi-circular or V-notches on both sides. These models are experimented with simple polariscope under tension, and it is explanined that K/K₀ is equal to P_r/P at γ≤0.6, and P_r/P can be calculated by brief calculation of the elementary theory.

View full abstract

This paper reports stress distributions in epoxy resin bars with reinforced circular holes under tension. The holes were reinforced by bonding aluminum rings to the bars inside the holes. The stress distributions were determined photoelastically. Various proportions of circular reinforcements and different widths of bars were investigated under pure tension. Stress distributions were determined in the epoxy resin part of the bar, and the maximum shear stresses were given special attention. The relation between the maximum shear stress and the width of the bar was determined, and the ranges for which the theoretical solution for an infinite bar approximates that for a finite bar were defined.

View full abstract

In the structural members under kinetic heating, it is usual that the deformation problems such as buckling are considered under uniform temperature distribution through the thickness and that the thermal stresses are simply given only by the temperature distribution without taking account of the deformation. However, if the temperature gradient exists through the thickness, the members may deflect from the beginning of heating, exhibiting no buckling behavior, and the thermal stresses will be thereafter made clear of itself. It is the purpose of this paper to present the above-mentioned fundamental purport of the problems under non-uniform temperature distribution in a simple thin beam having rectangular cross-section, not to make it ambiguous because of mathematical complexity, before entering into the plate problems. The two cases of clamped and simply supported ends where axial displacements are restrained, were discussed taking account of the finite deformation. When the temperature distribution exists along the length such as d²T₂/dx²≠0(T₂ is the temperature gradient through the thickness and x is the axial coordinate), the beam deflects from the beginning of heating. While, when d²T₂/dx²=0, the clamped beam does not deflect initially, even if T₂ exists along the beam, but the simply supported beam deflects initially, if T₂ exists at both ends. In other cases, it will be seen the deformation behavior such as in the Euler buckling.

View full abstract

The buckling of cylindrical shells under external pressure was analyzed on the basis of the theory of finite deformation. Restricting the authors' consideration to the case of boundary conditions, i.e.clamped straight edges and simply supported circumferential edges, the relations between the pressure and the deflection were obtained. The conclusions were as follows : 1.The "Durchschlag" phenomenon may take place more easily, if the values of 1/λ, 〓₀ and a/t are larger (λ=(y₀/l)², y₀=a〓₀, see Fig.2.). 2.The upper buckling pressure is lower, if the values of 1/λ, 1/〓₀ and a/t are larger. 3. The main characteristic for the case of cylindrical shell, which differs from that of spherical shell, lies in the fact that upper and lower buckling pressures have not minimum values against α(=(a<〓₀>²)/t). The cases of other boundary conditions can be analyzed with the same process as described in this paper.

View full abstract

In the present paper, the two-dimensional stress distribution in an orthogonally aeolotropic plate, bounded by two concentric circular contours, is treated. The same problems, in the case of isotropic material, have been solved easily by the use of the polar co-ordinates. In the case of aeolotropic plate, however, the solutions are rather complicated because of the directionality of its elastic properties, and the author treats it by the perturbation method using the orthogonal co-ordinates, similar to that of th previous papers. Numerical examples are worked out for three kinds of material, showing that considerably large stress concentration may occur in some aeolotropic plate even in the case of uniform inner pressure.

View full abstract

As to the plane stresses of orthotropic material, various studies have been worked out by many investigators, but the stresses in a plate bounded by lines asymmetric with respect to the elastic axes of the material, have been little known now. In this paper, some problems of this kind are treated by the use of Beltrami-Cayley's conformal mapping method. As examples, the stress distributions in (1) a semi-infinite plate, (2) an infinite plate with an elliptic hole, (3) an elliptic plate, which has boundary lines asymmetric with respect to the elastic axes of the material, are calculated. The results obtained show significant differences from these in isotropic plate and in orthotropic plate with symmetrical boundary to the elastic axes of the plate.

View full abstract

In this report, the buckling problem of a simply supported triangular plate having inner angles of 30, 60 and 90 degrees under uniformly distributed compressive and shearing load is solved. In the case of uniform compression, the exact solution is obtained as Eq. (11) by combining the solutions for a simply supported rectangular plate OACD in Fig.4. In the general cases as shown in Fig.1, the solution is obtained by the Galerkin's method. The numerical results obtained giving the bucking load are shown in Table 1, Figs. 5, 6, 7, 8 and Eq. (18).

View full abstract

In this report, the buckling problem of a simply supported equilateral triangular plate under uniformly distributed compressive and shearing load is solved by the same method as described in the previous paper. In the case of uniform compression, the exact solution is obtained as Eqs. (5) and (6), the former giving an asymmetrical buckling corrugation with respect to OH in Fig.1 and the latter a symmetrical one. The other solutions which are symmetrical and asymmetrical with respect to AH' and BH'' in Fig.1 are given as Eq.(9) by the combination of Eqs. (5) and (6). In the general cases as shown in Fig.1, the solution is obtained by the Galerkin's method. The numerical results obtained giving the buckling load are shown in Table 1, Figs.6, 7, 8, and Eq.(18).

View full abstract

In this paper, the strength and deformation of a curved plate cut out from a cylindrical shell and subjected to hydrostatic pressure, which has the maximum value along one curved edge and zero value along the other curved edge, is discussed utilizing a fundamental differential equation introduced by the author. The method of solving and results here obtained can be utilized to the design for curved sluice gates or arch dams.

View full abstract

Failure experiments on nine thin-walled cylinders whose lateral sides were subjected to uniform external pressure were conducted, for the purpose of finding the method of determining the initial out-of-roundness of actual cylinders to be used in the existing theoretical formulae for the failure pressure. The cylinders for the present experiments were of such dimensions that their computed values for yielding failure pressure were smaller than those for elastic buckling pressure, if the cylinders were completely round. The investigation revealed that : (1) The initial out-of-roundness need not be taken into consideration on the calculation of the failure pressure for the relatively round cylinders as the machined cylinders used in the present experiments. (2) It is appropriate to determine the initial out-of-roundness by a method similar to that suggested by Holt for the relatively unround cylinders as the plate-worked cylinders.

View full abstract

A yarn package may be considered as an elastic body consisting of concentric yarn layers around an axis of rotation, with different properties along tangential and radial directions. In this paper, stress functions of the anisotropic elastic body is applied to yarn package, and it is shown that there exist three regions in the yarn package, namely tensile stress region, compressive stress region and internal tensile stress region. The result of analysis matches well with the experimental results and this can be used advantageously for the control of tension force of yarn while building a package to avoid the bulge at the end of the package during rewinding.

View full abstract

It has been noticed in rotating bending fatigue test on 125φmm smooth surface specimens, described in part 1, that shell patterns appear on fractured surface when radical change of the stress is made on the specimen. The author has developed a general relationship between stress and propagation speed of fatigue cracks from analytical studies on shapes of the shell pattern, which may be summarized as follows : 1. The shell pattern shows a definite shape, which is drawn as a family of curves depending upon arc length of the crack. 2. Assuming that the stress σ (kg/mm²) on the edge line of the fractured surface is equal to the stress on a beam having a sectional shape same as that of the specimen having a reduced area by the fracture, σ was computed on a number of samples of shell patterns, as a result of which it has been found that the relation between the propagation speed v_c (μ/c) and σ may be obtained from the following formula : [numerical formula] Where : σ₀ is a critical value of σ;=11.5kg/mm². Where σ exceedsσ₀, the propagation of crack takes place. B is a constant depending upon the size of the specimen. B=17 for 125φmm specimen.

View full abstract

In this report, at first, the effect of the roughness of the testing surface of specimen on the Rockwell hardness was investigated, and then the effect of the roughness of the reverse surface was investigated, after the experiment on many specimens which were of different hardness or material. As a result, these effects on the Rockwell hardness number have been made clear, and it has been revealed that the roughness of the testing surface of specimen affects the hardness number, but on the other hand the roughness of the reverse surface does not influence the hardness number. The critical roughness curve which shows the relation between the surface condition ranging from the maximum (critical) roughness for the testing surface to free-from-the-effect of the roughness and the hardness of specimen was given. And after discussion of this result, it was presumed that the so-called "effect of the roughness of the reverse surface of specimen" which is told to be existed in industrial measuring is not due to the roughness itself but to the more direct causes related to the condition of the reverse surface. And also, the effects on three hardnesses, the Shore, Vickers and Rockwell, which are most widely used nowadays, were compared in this report.

View full abstract

The author depicted the allowable maximum external load diagrams and the varieties of the clearances between the most proper maximum clamping forces and the minimum clamping forces for various kinds of materials and dimensions under pulsating tension using the minimum values for some kinds of tensile strength picked up from JSME data sheets. Using these diagrams everybody will be able to decide the most proper dimensions, the most proper clamping forces and the maximum external loads adequate for the given bolt-connections.

View full abstract

[in Japanese]

View full abstract

Mechanisms of work hardening and recovery following an abrupt change in stress were discussed on the basis of the behaviour of dislocations. It was suggested that there would be a mechanism, which might act as a mechanism of recovery when the applied stress was abruptly decreased, while it might act as a mechanism of work hardening when the applied stress was suddenly increased. A mathematical expression of this mechanism was determined in a rather simpler form from the data of creep tests with sudden change in stress. Using this expression as well as other characteristic functions, which have been determined from the results of tension tests and creep tests with a duration of stress rest, it was also shown that the authors' theory of deformation well predicted the behaviour of mild steel after a sudden change in stress in primary or secondary stage of creep.

View full abstract

The shape of rubber vibration isolators are not limitted as simple bar type with circular or rectangular uniform cross section, but truncated cone or truncated pyramid type rubber and rubber bush are often used. So as the third item of shape effect of rubber vibration isolators, the author made some calculations and experiments, and discussed about their spring constants. Especially truncated rubbers are used as the elements of compound type isolators. Then calculations about such rubbers were made comparing with those that have uniform cross sections.

View full abstract

The author has found experimentally that sub-harmonic oscillations and "summed and differential harmonic" oscillations take place in a shaft system with non-linear spring characteristics and multiple degrees of freedom. In the present paper, these vibrations are treated mathematically and analytically. Conclusions arrived may be summarized as follows : 1) The equation of response curves of sub-harmonic oscillations of order 1/2 has qualitatively the same construction as that of rectilinear vibratory system. The hard or soft spring characteristic results in response curves of both hard and soft spring types. 2) When two vibrations taking place at the critical speed of "summed and differential harmonic" oscillations are both to forward or both to backward precessional whirling motions, only the summed harmonic type can occur and the differential type does not take place. When "summed and differential harmonic" oscillation consisits of forward and backward precessional motion, only differential type has possibility of occurrence.

View full abstract

For convenience' sake of designing the foundation and mounting of Hammer-Forging Machine, I have introduced the vibration formulae based on impact force, in case its shape is rectangular and saw teeth. These formulae would solve the probrem more accurately for actual phenomena than usual formulae which are introduced by the percussion theory on semi-rigid body. By strict consideration of physical meaning, some of these formulae are solved by conventional treatment which are introduced by the percussion theory on semi-rigid body. By strict consideration of physical meaning, some of these formulae are solved by conventional treatment and contain some approximate solution. But I have found that the result of the foundation and mounting practically designed for the 1/2 ton Steam Hammer and 1 ton one by these formulae and the various data, I have reported, are more similar to actual phenomena. I think, these formulae are convinient for service and also valuable for practical design.

View full abstract

In the first report the foundamental equation about characteristics of the tup motion was replaced with the algebraic simultaneous equations of higher order applying successive approximation of integral, and their approximate solution was obtained. Now in this case several numerical calculations for a 500 kg air hammer of Beche type are attempted through the solution as follows. Characteristics of the tup motion are to be influenced by the speed of crank shaft driving the pump piston and the most adaptive or the principal speed which means that one cycle of single blow accords with one cycle of pump piston, is to be found. Moreover the torque of crank shaft, the air pressure and temperature in both cylinders are calculated. And then referring to the desertation of experimental research "Untersuchung eines Schmiede-lufthammers" the writer has found that characteristic curves obtained by our analysis are similar to these experimental records researched through the reliable installations with pickup, indicator and recorder.

View full abstract

Servo-type analog-digital converter is a converter whose output (digital) is to follow the input (analog) on the same principle as the automatic balancing instrument works. In this converter, there are three principal components ; that is, counter conversion of digital to analog value (decoder), amplification of error between input and counter-converted analog value, and logical circuits which control the counting action of reversible counter according to the amplified voltage signal. The tentative converter is remarkable in three aspects. First, it has a stable AC amplifier working on the high carrier frequency transistor chopper. Second, output of amplifier is not demodulated to DC voltage, but is connected as it stands to the logical circuit, and so the response speed is very high. Third, decoder, using the MIT-type network, is a purely electronic network with transistors only as active elements.

View full abstract

The author studied the synthesis of automatic car-retarder speed control device. It is intended to make the actual exit speed of car coincide with the reference exit speed, and the controlled variable hereby is not the continuous function of time as it is usually. Cars, the controlled element, are replaced each time and the ratio of the weight of the heaviest car and the lightest car is about 6 : 1. The brake force reduces by half when the car is near inlet and outlet of the car-retarder. The author assumed two kinds of control methods ; the first is to regulate the car speed so that the speed-distance relation may duplicate the standard speed-distance curve ; and the second is to control the car speed by switching method. The former is classified into the continuous and the on-off control method ; and the latter into the one step and the two-step switching control method, and they are called Cases I, II, III and IV respectively. The author investigated the control process under the varied conditions in each case, calculated with electronic digital computer and reach the conclusion that it was possible to control the exit speed of car by the method called Modified Case IV, that is, a slightly modified version of Case IV, with high accuracy of control and relatively simplified construction of device.

View full abstract

For two systems of ACC having a constant pressure point at the main steam pipe, optimum controller settings are investigated. (Part I) System using only pressure and (Part II) System using pressure and difference between supplied heat and load. 25% damping is used for the optimum condition. For part I, expressing gain by k₁, and reset time by λ₁T_B, k₁, λ₁ and ρ=T_B/T_F have a certain relation, which is given in graph for ρ=5, 2.5 and 1.25. For part II, expressing gain of pressure by k₁, gain of difference between heat and load by k₂, k₂=k₁/5.41ρ-1 is held for P controller. Where T_B, and T_F are the time constants of thermal capacity and heat transfer respectively. Maximum pressure deviations for step changes of load are given in another graph. Giving required control characteristics, necessary system and controller settings can be obtained by the above graphs for given boiler parameters.

View full abstract