An analysis is presented to predict the deformation and distribution of stresses in a rotating disk with variable thickness and temperature gradient. The theory is based on the Mises criterion using the time hardening or the strain hardening rule. A result of experiments is presented. Spin tests were carried out on the specimens made of 13 Cr steel and of 300 mm outer diameter, 60 mm inner diameter, and 15 mm thickness. The deformation and distribution of the residual stresses were measured to investigate the change of stress distribution in the transient period. Fairly good agreement was found between theory and experiments.

View full abstract

In this paper experiments and theoretical considerations have been made on the burst strength of a high speed rotor which is used in an axial blower or gas turbine. According to the experiments the strength of a rotor is influenced by the centrifugal force induced by blades and the ductility of material of rotor ; the strength of a rotor rises above that of a rotating disc as the centrifugal force in a ductile material increases, though it doesn't rise in a small ductile material. The influences of the centrifugal force of blade and the ductility of material on the strength of a rotor are explained by the theoretically considerations and the results agree well with the experiments.

View full abstract

The fatigue tests on 0.23%C steel were carried out with two superimposed cyclic Ioadings with different repeating speeds, in the case when the repeating speed of the secondary stress, the fast stress wave σ_a2, was 2000c/min, and the primary speed, the slow one σ_a1, was 50 or 420c/min. The ratios of those stress amplitudes σ_a2/(σ_a1+σ_a2) were selected in the range from 0.75 to 0.27. Those complex stress patterns were analyzed by the peak pair count method, and the fatigue lives and endurance limits were calulated with the formula ; the method and the formula were proposed by the authors. The calculated values agreed well with the experimental results, even in the case where the primary stress amplitude was larger than the secondary one. It was also clarified that the peak pair count method is more widely applicable to those stress patterns than other count methods.

View full abstract

Tensile fatigue test of Cu-Al alloy (with 9% Al) in vacuum at high temperature was performed by the high-temperature vacuum tensile-fatigue tester made to work at pressure 10^-5mmHg, up to 700°C, under load of maximum 50kg and with load frequency 1∼10 sec^-1. And slip lines and fractured faces in fractured specimens were observed by optical microscope. The relationship between the present results and the results of tests on fatigue strength and high-temperature vacuum tensile-strength⁽⁸⁾ was discussed. The empirical formula for the relationship among the cyclic load ω, the number N^^- of load cycles to cause the fracture and testing temperature T (°K) can be given by N^^-=N^^-₀ (ω/W₀)^{-B^-} exp [-A' (T-T₀)²], where A'=0.88×10^-5°K^-2, B^^-=8, N^^-₀=4×10², T^^-₀=360°K and W₀=46.5kg.

View full abstract

In der vorliegenden Mitteilung wurde die VerschleiBgrenzschicht, die im Vorgang der Reibung bei Raumtemperatur auf der VerschleiBflache der Probe gebildet worden war, untersucht, wobei man von der Seite des Gasgehaltes und der AufschweiBung die Erfolge betrachtet hat. Dazu wurden einige Kohlenstoffstahle und Sonderstahle als die stillstehenden Proben, mit den die rotierende Probe aus dem Kohlenstoffstahl mit 0, 76% C gepaart wurden, gewahlt. Diese Paare wurden der VerschleiBprufung bei Raumtemperatur unterworfen. Nach der VerschleiBprufung wurde die VerschleiBflache einiger abgenutzten Proben mit dem Elektronenmikroskop und durch das Elektronenbeugungsverfahren untersucht. Weiter wurde der Fe-Gehalt in der VerschleiBgrenzschicht, die auf der zur VerschleiBflache bzw. Reibungsrichtung recht geschnittenen Flache bemerkt worden war, mit dem Elektronenstrahlmikroanalysator gemessen. Unter einigen Annahmen hat man von diesem Fe-Gehalt den Gasgehalt in derselben Schicht vermutet.

View full abstract

Though many methods have been proposed for calculating the strength of a car body, the energy method regarding the side frame as Vierendeel truss is employed by J.N.R. as its standard method. Here will be described the principle of the calculation and the program for the electronic digital computer, where the initial data (specific quantities of cross section, dimensions of various parts, elastic coefficients, condition of loading, etc.) are given as input ; and sectional forces (bending moment, shearing force and axial force), stress of various parts and deflection of nodal points of lower chord can be obtained as output.

View full abstract

By means of"eigen-value method"(which defines critical condition as the divergent point of flow equations), critical flow rates of two-phase fluid are predicted with"energy model", and are compared with the results obtained by existing methods (which define critical velocity as sonic velocity or derive it from dW/dp=0^* relation by the use of"entropy model"). And the following results are concluded ; (^*W : flow rate, p : pressure) (1) By the use of"eigen-value method", both"energy model"and"entropy model"can be treated easily by the existing method. (2)"Energy model"is not generally equal to"entropy model"and this difference comes from different transfer velocities of kinetic energy between the phases. (3) As for"energy model", the derivative of slip ratio (ds) can be neglected in evaluating critical flow rate, where an analytical expression of it is derived. These predicted values are satisfactorily compared with the existing empirical data.

View full abstract

Two-phase pressure drops have been measured in a vertical tube of inside diameter 24.5 mm in the liquid mass velocity range up to 3600kg/m²sec. The data are well correlated by the Lockhart-Martinelli method over the whole range of mass velocities. Martinelli's multiplier for the two-phase pressure drop, φ_L, does not vary with the total or liquid mass velocities in contrast with Petrick's results for rectangular channels. The results obtained were compared with other data and a fairly good agreement was found. The relation between a two-phase friction factor and Reynolds number, each based on the linear velocity of the liquid phase is almost identical to that for a single phase flow, i.e., the Blasius'equation, except in the case of very high air-velocities.

View full abstract

A theoretical formula for calculating the leakage is derived taking the carry-over of kinetic energy of the fluid from the preceding throttling into consideration. The carry-over factor in the formula is evaluated from the experimental results obtained by the radial labyrinth glands for compressed air. A consideration for the carry-over is presented assuming that the flow issuing from the throttling fin broadens out proportionally with the radial distance in the expansion chamber, and it is shown that the carry-over factor is a function of the ratio s/δ and the resistance coefficient of the flow that depends on the ratios h/s and δ/h, where s is the width of the chamber, h is the height of the fin and δ is the tip clearance of the fin. The results of the experiment performed to determine the discharge coefficient of a single radial fin are also presented for three distinct types of fin.

View full abstract

In the first part of this paper, the modes of vibration in the system with n parallel boiling channels and one downcomer and their stability are investigated. In the second part, a reverse flow, in which water flows downward in one or more heated channels, as if they were downcomers, and its stability are investigated. As the result, the following facts are found. (1) There are n normal modes of vibration in the system with n parallel boiling channels. One of them is an in-phase mode of vibration in which all channels vibrate in the same phase, and the other n-1 modes are mutual vibrations between channels, in which the sum of variations of flow rates is zero so that the downcomer gives no contributions to them. (2) The in-phase mode of vibration is more difficult to generate, as the number of channels becomes larger. Then, the mutual vibrations between channels, in which if flow rates are increased in some channels the same flow rates are decreased in other channels, are predominant in the system with many parallel boiling channels. (3) An orifice or throttle in the downcomer does not give the system any effects to damp the mutual vibrations between channels. (4) Reverse flow can be generated in the system with parallel boiling channels. The larger the flow resistance in the downcomer, the larger subcooling, the larger the difference of heat power between the channels, and the larger the number of channels is, the more easily a reverse flow is generated.

View full abstract

In order to justify the theoretical analysis on the hydrodynamic instability in boiling channels presented in previous papers, experiments are conducted using the systems with a single boiling channel, two parallel boiling channels and ten parallel boiling channels. The following facts which have been predicted theoretically are recognized experimentally. (1) The hydrodynamic instability in boiling channels is caused by the negative resistance due to the characteristic curves of pressure loss. Then, if subcooling becomes smaller, the system becomes stable. Exit void fraction has no direct influences on stability. There can be the case where the system becomes stable, when heat input is increased. The hydrodynamic oscillation can be generated also in the system without a riser. (2) Though, in a single boiling channel, a throttle in the downcomer gives damping effects to the system, the out-of-phase modes of vibration between channels can remain in parallel boiling channels, even when a throttle is included in the downcomer. Reverse flow can be generated in the system with parallel boiling channels. (3) As the number of boiling channels becomes larger, the in-phase-mode of vibration, in which all channels vibrate in the same phase, is more difficult to generate, and the mutual vibrations between the channels are predominant. (4) As the number of boiling channels becomes larger, there exist more equilibriums corresponding to reverse flow, then a reverse flow is generated more frequently in a multi-channel-system. In a multi-channel-system, there exist several channels which are settled in reverse flow.

View full abstract

Experiments of nucleate pool boiling were carried out on three kinds of heating surfaces and five kinds of liquid in the range of liquid level from 1mm to 30mm. As a result, the following facts were disclosed. It was found invariably in each liquid that coefficient of boiling heat transfer remains nearly constant at the liquid level higher than that which is inherent in each liquid and that it increases as the liquid level becomes lower than that value. The slope of boiling characteristic curve in such a low liquid level is different from that of usual liquid level. The effect of liquid level on boiling heat transfer coefficient depends on the liquid properties and heat flux and becomes more conspicuous as heat flux decreases. According to the dimensional analysis based on the facts mentioned above, coefficient of heat transfer was succesfully correlated.

View full abstract

A thin liquid layer and a dry part in a contact area of a boiling bubble on a wetted heating surface was clearly obserbed by experiment using an electro-conductive glass. The thickness of the liquid layer was calculated by reasonable assumption that the shape of a contact bubble on the heating surface is semi-sphere, etc. It was found to be a very thin layer (roughly 10 μ order). And then heat transfer through the thin liquid layer was calculated, and it was found that its coefficient of heat transfer in the contact area of the bubble was very good (roughly 10⁵kcal/m²hr°C order). Therefore we could understand that in the boiling heat transfer a main part is played by the thin liquid layer in the contact area of the bubble.

View full abstract

A dry part in a contact area of a boiling bubble on a wetted heating surface was observed by experiment using an electro-conductive glass. And size of the dry part was measured in the experiment, and it was compared with calculation based on vaporization by heat transfer through a thin liquid layer in the contact area of the bubble. And it was found that area of the dry part was roughly proportional to superheat of the heating surface and contact time of the bubble. And ratio of area of the part to total contact area in the heating surface was about 10% for any heat flux except very high and verly low heat flux at an atmospheric pressure in water of suturation.

View full abstract

In this paper, the authors investigated the bending fatigue properties and the behaviour of residual stress generated on the root fillet under the loading conditions corresponding to repeatedly reversed gears and idle gears, by using the bending fatigue testing machine of new type, which was developed by the authors. The main results obtained from this experiment are summarised as follows. (1) The bending fatigue limit of repeatedly reversed gears increases with the number of load cycles in one loading stage n. And there exists a linear relationship between σ_up"and log₁₀n. (2) The bending fatigue limit of idle gears decreases by about 40% compared with that in one side loading conditions. (3) In repeatedly reversed gears, compressive residual stress is generated on the root fillet of the load application side and tensile residual stress on that of the opposite side alternately, with the periodical change in the direction of load application. And in idle gears, compressive residual stresses are generated by alternating load on both sides of the root fillet.

View full abstract

Der vorliegende Aufsatz enthalt die Untersuchungen an Schraubenradgetrieben. Aus Versuchen wurden GesetzmaBigkeiten und Berechnungsgleichungen fur die Tragfahigkeit der Schraubenradgetriebe abgeleitet. Spezifischer VerschleiB, Grenzlast, Verlustleistung, Reibzahl und Rauheit der Zahnflanken wurden fur die verschiedenen Drehzahlen, Ole und Werkstoffpaarungen ermittelt. Die Grenzlast steigt mit der Drehzahl bei der Werkstoffpaarung Stahl/Bronze, sie fallt mit der Drehzahl bei Stahl/Stahl und sie andert sich nur wenig mit der Drehzahl bei Stahl/GrauguB. Die fur Schraubenradgetriebe ermittelte Grenzlast steht in Beziehung zu der im FZG-Stirnradtest fur die gleiche Werkstoff-und Olpaarung ermittelten Grenzlast. Fur die Vergleichsrechnung sind die Hertz'sche Pressung und die Gleitgeschwindigkeit am Zahnkopf maBgebend. Zum Berechnen bei der beiden GroBen sind neue Gleichungen angegeben.

View full abstract

Mating of a gear-tooth is intermittent, and various factors, which affect the lubrication characteristic, vary quickly during a tooth mating. Therefore, it is very important to analyze the dynamic phenomena of oil film in order to clarify the mechanism of gear lubrication. The hydro-dynamic squeeze-effect was introduced into the lubrication theory, and the transient performance of oil film caused by discontinuous variation of contact state or load was analyzed. Experimentally, the transient thickness of oil film was measured by using special model disks, and it was found that the results were qualitatively in good agreement with the theoretical results. It is deduced that the dynamic phenomena of oil film can be fundamentally explained by the dynamic lubrication theory into which the squeeze-effect is introduced.

View full abstract

Oil film state in mating gears is considered to be quite a dynamic phenomenon, and it has become very important to analyze the dynamic performance of oil film between mating gearteeth. The hydrodynamic squeeze-effect and the tooth deflection were introduced into the lubrication theory, and the variation of oil film thickness between mating teeth was calculated. Experimentally, the variation of oil film thickness was measured by the electrical resistance method from the beginning to the ending of a tooth mating, using specially manufactured test gears which do not form a parallel resistance circuit even in mating of two pairs of teeth. The theoretical result is in good agreement with the experimental results particularly at the beginning part of mating. It is clarified that the principle of oil film formation on mating tooth surfaces can be fundamentally explained by the dynamic lubrication theory which is introduced in this paper.

View full abstract

Gears with circular-arc tooth-curve can easily be ground with the cup-shaped grinding wheel and a desirable crowning can readily be given to the tooth surface. The gear grinding machine can be of the simplest construction. Further, it can grind parallel-depth toothed spiral bevel-and hypoid-gears theoretically correctly with great ease. Therefore, high precision, good tooth-bearing gears can be made at the first setting without any trial cutting. The parallel-depth tooth system should be adopted especially when small numbers of various gears are to be produced. However, since the cup-shaped grinding wheel and the tooth surface being ground contact as if to envelop each other, the arc of contact of the abrasive grain is long and the cooling and flushing effects of the grinding fluid cannot be achieved sufficiently. Consequently, there is a fear of burning of the tooth surface. By the addition of a small revolving motion (r=0.5 mm, n=60 rpm) to the shaft of the grinding wheel, intermittent grinding is effected, where the grinding fluid flushes and cools the grinding region sufficiently, and the grinding horsepower decreases remarkably because of the decrease of grinding friction. Thus, the grinding efficiency is markedly increased. Even the plunge grinding, or the so-called Formate Grinding, can be performed successfully.

View full abstract

The influence of dressing condition of a grinding wheel on the grinding fluid performance was investigated experimentally in free infeed plunge grinding. The results can be summarized as follows : The effects of the emulsifier of additive concentrations in emulsion type grinding fluids on the rate of stock removal and the grinding ratio vary markedly with the dressing conditions such as dressing speed and sharpness of the dresser, which govern the character of the dressed wheel surface. Every grinding fluid has an optimum fluid concentration for the best grindability obtainable with a definite grinding wheel, the optimum concentration being shifted to a higher value by increasing the dressing speed or sharpness of the dresser. When free infeed plunge grinding is applied for the evaluation of the performances of fluids, the test results should be examined in due consideration of the dressing conditions.

View full abstract