The impulsive pressure is detected by the local pressure sensor, used in the previous report, at the 2nd collapse-reexpansion stage of a spark-induced cavitation bubble near a solid boundary. At the same time a flash photograph of the cavitation bubble is taken, and it is revealed that a microjet might be generated at the 2nd collapse-reexpansion stage of the cavitation bubble and the impulsive pressure of the microjet is about 1-2MPa, when the cavitation bubble is generated at 7-8 mm from the solid boundary. It is also revealed that the time lag between the microjet impact to the solid boundary and the shock wave generation is about 40-60 μs.

抄録全体を表示

Distributions of particle velocity and concentration were investigated in a pipeline with a 90° horizontal-to-vertical bend. Measurements were performed with an optical fiber probe for three characteristic flow patterns, i. e. dispersed flow, dispersed biasing flow and sliding bed flow, as classified in our previous work. It was found that the distributions were strongly affected by the flow patterns. The velocity distributions in the sliding bed flow were nonuniform and had large gradients, while the distributions in the dispersed flow were nearly uniform across the section. The difference due to the flow patterns was more marked in the results for concentrations. The profiles of concentrations were extremely distorted in the bend owing to the centrifugal force. Moreover, unexpected results were obtained in regard to the variation of the concentration profiles after the bend. At first, the profiles showed an opposite asymmetry in the straight pipe, after which they gradually became uniform. In spite of the above distorted distributions, calculations of particle velocities based on the one-dimensional model agreed well with the experimental results in the straight pipe after the bend.

抄録全体を表示

The trajectories of small particles in a bifurcating laminar flow are studied by numerically solving the equations governing the viscous fluid and the particle. The following assumptions are introduced: The flow is incompressible, two-dimensional and steady; the shape of particle is spherical; the interaction between particles is negligible; the effect of gravity is ignored and the interaction between the particle and the fluid is governed by Stokes' drag law. The motion of the particle is governed by the Stokes number, the Reynolds number, the ratio of flow rates and that of the widths of the branches. After a very long calculation, the separating lines, by which the particles in the upstream channel are separated into particles leaving for the straight and for the side branch, are found numerically. The location of the separating line is found to depend strongly on the ratio of flow rates in the two branches, but the effect of the Reynolds number on the separating line is comparatively small. Collisions between the particles and the channel almost arise on the wall in the side branch. The distribution of collision frequency along the wall of the side branch is calculated and found to take the maximum value at a Stokes number.

抄録全体を表示

When alternating magnetic field is applied to magnetic fluid in a vessel, unsteady interfacial instability occurs on the free surface. In the paper, this kind of instability is investigated theoretically on the basis of the inviscid non-polar theory of ferrohydrodynamics. The analytical solutions of the shape of free surface, the induced magnetic field and the velocity distribution are derived, and it is clarified that the instability is due to a parametric reasonance governed by Mathieu equation. Neutral stability curves are obtained for given wave numbers by solving the equation. As the frequency of the applied magnetic field becomes closer to the natural frequency of the fluid in the vessel, the instability tends to occur more easily. Moreover, it is found that another steady instability can take place, if the wave length of the disturbance is short enough.

抄録全体を表示

Two-dimensional flows over a backward facing step of viscoelastic fluids have been solved by the finite difference method. The Maxwell model and Denn model are used as constitutive equations. The result has shown that behavior of a recirculating flow in the rear of the step is dependent of the Reynolds number (R), Weissenberg number (W) and rheological constants (n and s) of viscoelastic fluids. Recovery of velocity profiles in the downstream of the step is slower in viscoelastic fluids than in Newtonian fluids at the same Reynolds number. Using the present numerical method, it is possible to calculate the flow field up to R=100 and W=1.0.

抄録全体を表示

Impingement of an axially-symmetric, underexpanded sonic jet on a perpendicular flat plate is studied using flow visualization and a numerical technique. The numerical results which are obtained by a two-step Lax-Wendroff scheme under the assumption of inviscid flow are in good agreement with the experimental results. It is found that the maximum pressure is not always at the center on the plate, and that there may be a reverse flow in the central area on the plate.

抄録全体を表示

The present paper deals with the problem of a turbulent round jet discharged into a coaxialdeadend pipe associated with that of flow in a flue. This was already studied by several investigators and they found an interesting fact: depending on the geometrical ratio of the jet and the pipe, the jet reaches close to the bottom of the pipe in some cases and reverses midway far from the bottom to the opposite open end in other cases. To clarify this geometrical ratio governing flow patterns, some experiments and measurements are carried out for several ratios. Numerical computations are also carried out to solve the equations for vorticity and stream function by making use of the k-ε turbulence model. Finally, it is shown that the experimental results can be reproduced by the computations and the effect of the geometrical ratio is made clear.

抄録全体を表示

We have observed the interaction between a shock wave and an unsteady jet in a shock tube. From the results of the experiment, we have reported that the interaction between the shock wave and the unsteady jet in a shock tube was very strong, and it temporarily generated a hemispherically reflected shock wave. We could not clarify the cause of the interaction because of the limits of the experiment. Therefore, we used numerical computation to clarify it. The computations were carried out by solving two-dimensional Eulertype equations by means of the FLIC method. The results obtained from the computation generally supported those expected from the experiment. The results are summarized as follows : (1) The interaction between a reflected shock wave and the jet was very strong and it temporarily generated a hemispherical reflected shock wave. (2) The pressure, density and temperature of the apex of the hemispherical reflected shock wave were estimated to be 2 to 3 times higher temporarily, than those of the reflected shock wave in the absence of the jet.

抄録全体を表示

The effects of pH and Cl ion concentration on erosion-corrosion of mild steel in aqueous solutions were investigated using an impingement jet flow apparatus. It was found that the mass loss in ion-exchanged water decreased with jet velocity, and the mass loss in the water with 6.7 ppmCl reached a peak at 0.1 m/s. Such a peak appeared in the case of the other solutions, but the peak value depended on pH, and was absent when pH > 6.5, because the increase of pH suppressed the breakdown effect of the Cl ion. A large number of pits were generated on the mild steel subjected to the solutions with Cl ions, probably because the Cl ions caused breakdown of parts of the protective film.

抄録全体を表示

A domain decomposition method is applied to the solution of unsteady Navier-Stokes equations in primitive variable formulations for incompressible viscous flows in a square cavity and around a rectangular cylinder. In the present approach, a computational domain is divided into several subdomains, thus making the grid generations and flow computation much easier than for a single domain, especially in the case of very complicated geometry. The flow equations are solved by the method of lines using the rational Runge-Kutta time integration scheme, while the Poisson equations are solved by the SOR method. It is concluded that the present method essentially agrees well with methods using a single domain.

抄録全体を表示

Three-dimensional laminar flows in various curved pipes # (i. e., R_c/a=2 and 9 for curved 180° pipes and R_c/a=2 for curved 90° pipes are numerically simulated through use of the full Navier-Stokes equations. The boundary-fitted coordinate system is used in order to treat complicated boundary configurations like a strongly curved pipe. The obtained flow patterns and the magnitude of the secondary flow in the case of R_c/a=9 for a curved 180° pipe are in good agreement with the experimental results. The elliptic nature of the basis equation is very important in simulating these recirculating flows, especially in the case of small curvature. Two separation regions, one occurring on the outside wall near the inlet and the other occurring on the inside wall at the outlet, are found in the case of R_c/a=2. Not only is the crossover point between the shear stress maxima on the inside and outside of the pipe independent of the curvature and the bending angle of the pipe, but it is also independent of the initial profile at the straight pipe.

抄録全体を表示

The two-dimensional Euler equations have been solved by a finite volume method (FVM). The computational domain is divided into either quadrilateral or triangular cells. Spatial discretization of the Euler equations on each cell yields a system of ordinary differential equations (ODEs) in time. A rational Runge-Kutta (RRK) scheme is used to integrate the system of ODEs. The implicit residual averaging (IRA) and local time step have been employed to accelerate the convergence to a steady state. Numerical solutions are presented for flows past a circular cylinder at Mach 0.38, the NACA 0012 airfoil at Mach 0.8 and an angle of attack of 1.25°, and a channel flow at Mach 0.675. The effects of the shape of control volume and convegence acceleration techniques are investigated by comparing the results with those of the conventional finite difference discretization method (FDM). It is concluded that accuracy and efficiency of the FVM are comparable to those of the FDM.

抄録全体を表示

The flows around a fast-back-type car were investigated by the numerical simulations with k-ε turbulence model. The highly efficient calculation code on the three-dimensional and general co-ordinate system was developed for this research based on the finite volume method. The comparison of the calculation results with the experimental data showed that these simulations can predict the critical slant angle of the rear window. The effect on the grid size was also estimated by several examples.

抄録全体を表示

A new nondimensional parameter is proposed in order to estimate the two-dimensional characteristic values of a lifting cylinder with tangential blowing. Each experimental relation, between various two-dimensional characteristic values and the new parameter, is presented. In each case, a single approximate curve, as a function of this parameter only, is given. Note that he curves are independent of the blowing position and the blowing intensity. Then, a new method for estimating the two-dimensional characteristics of the cylinder by giving only the blowing position and the blowing intensity is described. In the case where the blowing position is in the separated region, this method can only be applied to a lifting cylinder which is at the condition of the forced reattachment of the separated shear layer. The results estimated by this method were fairly consistent with the experimental ones.

抄録全体を表示

The stabilizing structure for the phenomenon where the flow rate of a loop with a cross pipe switches nonperiodically between a high level and a low level under certain flow conditions was analytically and experimentally investigated. By numerical analysis, a pair of vortices appeared downstream from the point where the axes of the main pipe and the side branches crossed. As the upstream vortex grew larger, it drifted further upstream and swirl flows with a high pressure loss resulted. Conversely, when the downstream vortex grew larger, the flow stage became nonswirling with a low pressure loss. Two methods for stabilizing nonswirling flow were confirmed by experimentation. One is the use of a flow diverter and the other is the removal of the region where vortices appear by narrowing the downstream area at the cross pipe.

抄録全体を表示

This paper presents an experimental study on the flow around a stepped circular cylinder and the structure of the shedding vortices. The surface pressure distributions on the cylinder and the spectra of the wakes were measured for diameter ratios of 0.4 and 0.8. The main results are as follows. (1) The back pressures of the two stepped circular cylinders are higher than those of a straight circular cylinder. Therefore, the local pressure drag coefficients of the stepped cylinders are lower than those of the straight cylinder. (2) The structure of the shedding vortices is different between the diameter ratio of 0.4 and 0.8, being complex in the case of 0.4 and simple for 0.8. (3) In the case of 0.8, the signal representing the vortices can be expressed as a linear combination of two sine waves and the structure of the filaments obtained from it agrees with the result of a flow visualization.

抄録全体を表示

To estimate the usefulness of the wide-spacing approximation proposed by Ohkusu, Srokosz et al, the present paper treats the interaction of two floating bodies. First, the accuracy of the wide-spacing approximation is discussed by using the Green function by John and it is shown that the accuracy of this approximation is of order exp (-l) where l is the distance between two bodies. Second, the numerical calculation is carried out by using the boundary element method about the two-dimensional rectangular floating bodies, and its results are compared with the wide-spacing approximation. It is shown that the wide-spacing approximation is very accurate in the heave motion, but the accuracy in the sway motion is not as good in the case where the distance between bodies is less than 1.5 times the body width. Finally, the energy absorption coefficient in two heaving rectangular bodies is examined by using the wide-spacing approximation.

抄録全体を表示

The conventional assumption for triple products, i. e., the gradient-type diffusion, is not correct in near-wall turbulence and thus the use of this assumption in Reynolds-stress and heat-flux equation modelling may introduce serious errors into the calculations. In the present study, first we have developed a statistical model for triple products of velocity and temperature using the knowledge of the probability densities and internal structure. Next, we have combined this model with the transport equations of triple products. It is shown that the proposed model works quite well in various flow fields.

抄録全体を表示

A study has been carried out experimentally as well as theoretically in order to understand the mechanism of stall margin improvement of casing treatment. A low-speed compressor equipped with a single rotor was used for the detailed measurement of the circumferentially-averaged-internal flow with and without casing treatment. Based on the experiment, the effect of casing treatment on total-pressure-loss characteristics of the blade row was discussed, and the result was incorporated into the theoretical analysis. Two-dimensional theory of the rotating stall, to predict the onset condition, was applied to each radial section in a stripwise method, and it was found that the agreement between the experiment and the theory was very good for all configurations, with and without casing treatment.

抄録全体を表示

An experimental study has been carried out to clarify the mechanism of stall margin improvement of casing treatment. The previous study revealed that the two-dimensional theory combined with experimentally obtained total-pressure-loss characteristics can predict with accuracy the onset of rotating stall for all compressor arrangements with and without a few types of casing treatment. In this study, two low-speed compressors were used to measure total pressure distributions down-stream of the rotor as well as within the blade passage. The behavior of the low-total-pressure fluid originating from the tip clearance leakage and the high-total-pressure fluid ejected from the treatment slots are discussed. Based on the experiment, a flow mechanism in which the casing treatment improves the total-pressure-loss characteristics of the rotors and thus improves the stall margin of the compressor was proposed.

抄録全体を表示

The problem of fluid film squeezing between convex and concave walls of the same curvature is studied theoretically. Their shapes are arbitrary but axisymmetric, and they are installed coaxially. One wall is fixed and the other is moving along its axis with an arbitrary velocity. Assuming that the ratio of the gap between the two walls to their radii is small, an approximate solution of Navier-Stokes equation is found in the form of the power series of the ratio with precision up to the third order. In order to confirm the accuracy of the approximate solution, it is compared with a numerical solution for the case of flat walls. They agree with each other. It is found that the results obtained numerically by other workers can be derived easily from the approximate solution. The effect of the curvature of the walls on the load capacity is investigated using sinusoidal and parabolic shapes for the walls. For the case of the moving wall with a constant velocity, the leading term of its effect is proportional to the second order of the ratio of the gap to the radius. For the case of the oscillating wall, the leading term of its effect on the mean load capacity is proportional to the third order. Their proportional constants are expressed in the form of a quadratic expression of the maximum height in the cross-sectional profile of the curved wall. The effect of the curvature for the convex wall differs in magnitude from that the concave wall.

抄録全体を表示

By means of a small pressure sensor built into the tooth root on the discharge side of the female rotor, the pressure change in the groove can be measured from the midpoint of the suction stroke to the completion of the discharge. Indicator diagrams were collected for a single-stage cycle and the indicated horsepower was calculated. In this manner, we could determine the gas compression conditions inside the groove the thus the data could be analyzed for furthering the studies on high-efficiency profiles.

抄録全体を表示

The effects of three design parameters, the number of disks, the clearance between casing wall and front shroud and the distance of two disks on fluid dynamic characteristics, of a laminar flow fan were experimentally investigated over a wide range of fan flow rates, and were discussed in relation to the measured velocity field at the outlet of the rotor. The pressure head was analyti-cally derived, and the results show that the fluid dynamic characteristics were improved by in-creasing the number of disks and by reducing the clearance between casing wall and the front shroud. The agreement between the predicted and the experimental results of the pressure head is satisfactory if the experimentally determined value of K_m was used.

抄録全体を表示

Velocity distribution-, turbulence shear-stress distribution and static pressure of the turbulent flow through a conical diffuser were numerically calculated and compared with the measured data. Numerical calculation was achieved by use of the momentum equation, the zero equation model for turbulence shear-stress and the integrated turbulence-energy production. The results mean better reliability of this method.

抄録全体を表示

Flow measurements were made in a shrouded centrifugal impeller, with much the same passage as the unshrouded impeller in the first report. A measured flow rate corresponded to a nearly zero incidence at the inlet. Using a five-holed pressure probe, the relative velocities and the reduced static pressures were measured on nine sectional planes within the passage of a shrouded impeller running at 1000 rpm. Influences of the secondary flow and the turbulent mixing on the wake were discussed, and the flows in this impeller passage were compared with that of the unshrouded impeller in the first report. The wake in the shrouded impeller occurs at the suction-side/shroud corner in the mid-passage of the impeller ; this region develops near the suction surface along the passage. The position of this wake differs from the region shown in the unshrouded impeller passage of the first report.

抄録全体を表示

Effects of an inlet swirl on pump performance and static pressure were investigated experimentally. When a swirling flow was introduced to a centrifugal pump, the pump head was increased/decreased depending on the direction of the inlet swirl, and efficiency deteriorated because of the increase in the hydraulic loss at he impeller inlet. The instantaneous pressures on the blade surface were measured by transducers mounted on the impeller, and the rapid change of static pressure caused by the interaction of the vortex core and the impeller blades was clarified. The static pressure change was found to be dominant near the leading edge of the blade, but decreased along the blade. Introducing a small quantity of air into the vortex core made it possible to visualize the behavior of the vortex core at the inlet of the impeller passages ; it was found that the behavior of the vortex core was quite different with the direction of the swirl.

抄録全体を表示

The equations of motion of a conventional piston in axial piston pumps and motors including metallic contact are derived. For swash plate types, the fluid film shape between the piston and the cylinder wall, contact forces and loss of power are numerically obtained in relation to both the operating conditions and piston dimensions. Especially, the length of the sealing part composed of the piston and the cylinder wall is shown to affect considerably the power loss. Finally, the validity of the analysis is discussed by the comparison with the results of a performance test of a pump for high water content fluid.

抄録全体を表示

An experimental study was performed for natural convection in a hemisphere heated from below. The purpose of the present study was to obtain heat transfer characteristics for a wide range of Rayleigh numbers and to clarify the behavior of thermal plumes in connection with the laminar-turbulent transition of such a flow. Heat-transfer and flow-visualization experiments were carried out using various liquids. The ranges of Rayleigh and Prandtl numbers for the experiments were 1×10⁶<Ra<5×10¹⁰ and 6<Pr<13 000 for the heat-transfer experiment and 5×10⁵<Ra<2×10⁹ and 6<Pr<13 000 for the flow-visualization experiment. Empirically obtained heat-transfer correlations and the behavior of the thermal plumes generated on the hot surface indicated that the laminar-turbulence transition occurs at Ra 10⁹.

抄録全体を表示

A numerical analysis by use of the k-ε turbulence model was performed on the flow and heat transfer characteristics of the strongly heated annular gas flow. Some modifications were added to the Nagano & Hishida's original k-ε turbulence model, in order to improve the prediction efficiency in the low Reynolds number region. This modified version is superior to the original one in that it predicts the high flux heat transfer performance more accurately than the original one does. Calculated radial profiles of the turbulence energy and of the average streamwise velocity suggest that no definite laminarization occurs, even in the highest heat flux case. It is concluded, therefore, that the annular gas flows are less vulnerable to the laminarization than the circular tube flows.

抄録全体を表示

The aim of this paper is to reveal the laminarization phenomena of the circular tube flows through numerical simulation, where the author's modified k-ε turbulence model was used. The second purpose is to confirm the ability of this model, which was partly established in the preceding report, for the annular tube flows. It was found that this model achieves a sufficient degree of accuracy in reproducing the streamwise variation of the Stanton number, which has not been obtained by other k-ε models. The turbulence energy profiles of laminarizing flows were clearly distinguished from case where the Nusselt number decreases simply because of the local change in the viscosity. Obtained criteria for occurrence of the laminarization were in good agreement with existing experimental results. It was concluded, therefore, that this model can be applied to various situations of flow and heat transfer.

抄録全体を表示

Heat transfer coefficients were measured for horizontal tube bundles in the freeboard region above an atmospheric fluidized bed of silica sand of 0.86 mm. The average heat transfer coefficients for each tube were correlated well by a simple equation including the effects of static bed height. Comparisons with the previous correlations demonstrate that the present form gives a more accurate and useful approximation for the measured heat transfer rates in the freeboard.

抄録全体を表示

An experimental study on heat transfer of a rectangular cylinder in the turbulent boundary layer on a plate was performed. The velocity of undisturbed flow ranged from 4 to 18 m/s. The height of the cylinder was 10 and 20 mm, and the width was 30 mm. The turbulent boundary layer thickness without the cylinder was about 14 mm. The local and mean heat transfer coefficients on each face were measured and discussed in relation to the fluid flow. The heat transfer performance was compared with that of the same cylinder in a uniform flow.

抄録全体を表示

This report deals with the experimental study which focused on the relationship between the local heat transfer coefficient of ribbed surfaces and the heat flux ratio (the ratio of heat flux on the ribbed surface to the one on the base surface). Experiments were carried out with a parallel channel with a ribbed lower surface and a smooth upper surface. Square bars were attached to the lower surface. Heat flux ratio ranged from 0.5 to 2.5, Reynolds number from 3×10³-1×10⁵, the ratio of pitch to the height of the rib from 2.5 to 60 and the ratio of the height of rib to the hydraulic diameter of the parallel channel from 0.032 4 to 0.174.

抄録全体を表示

The temperature distribution of IC packages on a printed wiring board is calculated by considering convective heat transfer by superposing the air temperature rise behind upstream packages. Conduction to the board and radiation to the surroundings are also considered. Two calculation methods are presented : An analytical method for the case when all packages have the same geometry and are placed with regular pitches, and a numerical method for the case when packages have arbitrary sizes, or are placed at arbitrary locations. Calculated temperature distributions agree well with the experimental results.

抄録全体を表示

It is very important to utilize the local cold energy in the winter season. In the previous paper, the authors suggested a house for large-scale and long-term storage of the natural cold energy such as natural ice or snow, and by theoretical investigation, clarified that this type of storage house is of benefit to users of cold thermal energy in the summer season. In this paper, theoretical results of the previous paper are verified by a practical experiment with a scale model. A storage house used for experiment is buried in the ground except for the roof, and is constructed with concrete, insulators and so on. The measured values of the melting rate of ice, soil temperatures and room temperatures agree well with values estimated by theoretical analysis.

抄録全体を表示

Condensation heat transfer of a nonazeotropic binary R113-R114 vapor mixture in a vertical tube is investigated experimentally and theoretically, for cases of both constant and varying wall temperature along a flow direction. Experimental results show that a minimum temperature difference is required for complete condensation at an outlet of the condenser tube. This temperature difference and the heat transfer coefficients can be accurately predicted by a theory which assumes a well-mixed condition in the condensate, that is, a uniform concentration profile in the condensed liquid. When the fin height is 0.2 mm, augmentation of heat transfer by an inner finned surface occurrs in the liquid film but not in a vapor phase. For both cases of the constant and varying wall temperature, the heat transfer coefficients correlate well when an averaged temperature difference between the wall and the bulk flow is used as a characteristic temperature difference.

抄録全体を表示

The characteristics of forced convective condensation of a binary vapor mixture on vertical and horizontal tubes were experimentally investigated using a R113-R114 mixture for vertical and horizontal vapor flows. A smooth tube and two kinds of finned tubes were used as test condensing tubes. It was concluded from the experimental results that the thermal resistance of a vapor diffusion layer is affected markedly by the turbulence generated by a wavy motion of the condensate liquid film and the fins of the condensing tube. The dependency of he heat-transfer coefficient on the vapor-flow direction and the fin shape of the condensing tube can be explained clearly by considering these effects of turbulence. A prediction equation for the heat-transfer coefficient which is derived from an analogy between heat and mass transfer with consideration to the effects of the turbulence generated at the tube surface can accurately estimate the experimental results.

抄録全体を表示

The contact area of boiling bubbles on the horizontal heating surface of transparent glass was observed from the back side of the heating surface through transparent fused salt in transition boiling. The boiling liquid for the experiment was Freon R113, and the experiment was performed under the saturation condition at atmospheric pressure. The behavior of bubbles and liquid on the heating surface was observed in transition boiling by means of photography. As a result, the ratio of the area of the microlayer in contact bubbles on the heating surface to the total are of the heating surface was decreased and the ratio of dry contact area of the bubbles on the heating surface was increased with increasing of the superheat of the heating surface.

抄録全体を表示

Vortical structures in a cross-shaped jet have been investigated, using a flow visualization technique by laser-fluorescent dye while controlling vortex evolutions under excitation. The corner regions of vortical structures close to the orifice undergo severe deformation due to the self-induced velocity ; the outward corners become flat and the inward corners generate outward ejection. The excitation at low frequencies produces the large-scale vortical structures, the strong self-induced velocity switching over the normal and diagonal axes of the cross-shaped vortices. The deformed vortical structures are stretched under the effect of the mean flow, discharging small vortices from the tips of corners through the cut-and-connect mechanism. The vortex stretching and the splitting to small vortices in the cross-shaped vortical structures are expected to enhance mixing significantly.

抄録全体を表示

The present study describes the experimental results on the characteristics of he impingement heat transfer caused by three slot jets. Experimental values were obtained for the dimensionless space H=0.5-3, the dimensionless pitch P=6-16 and the Reynolds number Re=500-8000. For laminar impinging flow, they were compared with numerical results. For turbulent impinging flow, two peaks of the local Nusselt number were obtained behind the second nozzle. The position of he second peak approached the nozzle with the decrease of the space between nozzle and impinged surface. The average Nusselt number between the central and the second nozzle was determined by the ratio P/H and the Reynolds number based on the pitch of the nozzles.

抄録全体を表示

Characteristics of critical two-phase flow through a very narrow channel and a labyrinth channel have been investigated. The 0.1 mm high channel was made of transparent glass plates to allow observation of the flow pattern. The critical flow rate was measured for various inlet conditions. For the rectangular channel, the measured flow rate under subcooled inlet conditions was larger than predicted by conventional models because of significant thermal nonequilibrium between phases. A new model incorporating nonequilibrium effects was propose, and the results were compared with experimental data. For the labyrinth channel, the measured flow rate was smaller compared with that for rectangular channel due to the relaxation of thermal nonequlibrium in the cavity. The critical flow rate in the labyrinth channel was also predicted theoretically.

抄録全体を表示

The laminar burning velocity and the turbulent burning velocity of propane-air homogeneous mixtures in a closed vessel were successfully measured from the difference between flame propagation speed, W_f, and unburned gas velocity, U_l, ahead of a flame front. U_l was measured with LDA, and W_f was determined by measuring the time of flame travel over a fixed distance. The flame front was detected by the deflection of a laser beam owing to the rapid change in the refractive index at the flame front. It was found that these measured values were reasonable as compared with the values determined from the configuration of flame and the combustion pressure.

抄録全体を表示

The following three methods of ignition delay measurements, ultraviolet radiation of OH, visible and ultrared radiation and pressure rise due to ignition, were experimentally compared in the spontaneous ignition of fuel sprays injected into a steady hot-air stream. The following results were obtained. (1) Ignition delays measured by the above three methods show a similar trend to each other in the Arrhenius co-ordinates, but the values are different and become larger in order for OH radiation, visible and ultrared radiation and the pressure rise. In particular, the difference is larger for a high air temperature or short ignition delay. (2) Ignition delay measured with visible and ultrared radiation is strongly influenced by soot formation. This suggests that the detection of visible and ultrared radiation may possibly be incorrect in some conditions of ignition delay experiments because the soot formation drastically changes in flow fields due to fuels or mixing condition.

抄録全体を表示

Vibrational Raman spectra of acetylene ν₂ band were theoretically obtained for the experimental use in a flame. The present simulation includes the effects of Fermi resonance and vibrational l-type resonance-doubling to achieve a more accurate evaluation of temperature and concentration. The spectra such obtained were compared with those measured using an excimer-dye laser system. The theoretical prediction appears usable in a relatively wide range of temperatures due to its high sensitivity to temperature change.

抄録全体を表示

The new empirical equations to represent the Sauter mean diameter of a spray injected from a diesel nozzle were reported in this paper. In order to determine the new equations of the mean diameter, drop sizes of a diesel spray were analyzed by a laser diffraction technique. Liquids with different viscosities and different surface tensions were tested. The maximum injection pressure and the maximum ambient pressure were 90 MPa and 3 MPa, respectively. According to the tendencies of break-up length and drop sizes of a spray for an injection velocity and an ambient pressure, the spray formation mechanism could be divided into two categories as a function of the injection velocity and physical properties of the liquid. Finally, dimensionless analysis for each category of a spray could lead tot the empirical equations of the Sauter mean diameter.

抄録全体を表示

In order to improve the accuracy of the numerical prediction of gas flow in intake ports, a porosity approach has been applied to control volumes adjacent to walls. Numerical prediction of intake port flow has been carried out under steady flow conditions. By comparing the results of numerical predictions using the porosity approach with those using the previous approach of representing wall surfaces with staircaselike boundaries, it was confirmed that the accuracy of flow prediction was improved considerably by using the porosity approach. It has therefore become feasible to predict pressure losses in intake ports quantitatively. Consequently, an improvement in accuracy may be expected in the prediction of volumetric efficiency and swirl intensity under actual engine conditions by applying the porosity approach.

抄録全体を表示

It is considered that the oxidation of soot particulates in the combustion chamber of diesel engines is influenced by their microcrystal structure. The microcrystal structure of a soot particulate is not clear however, and this investigation attempted to determine the microcrystal structure of soot particulates in the combustion chamber and exhaust by means of a high-resolution electron microscope. The results were as follows : the typical layer plane structure which characterizes graphite carbon is not observed in the exhaust of diesel engines, but some particulates display a somewhat similar layer plane structure. The structure of soot particulates is turbostratic because the electron diffraction patterns of soot particulates are patterns of polycrystals. The structure of soot particulates in the combustion chamber is similar to that of exhaust soot particulates. In the initial stage of combustion, a similar layer plane structure is observed, but after the heat release there is no clear layer plane structure.

抄録全体を表示

The heat transfer properties of various matrices for Stirling engine regenerators were evaluated and compared without mounting the matrices on actual engines, using a special test rig whith two rotary valves. On the test rig, a simple test piece of such a matrix is mounted and the two rotary valves allow heated air and unheated air to flow into the test matrix alternately in opposite directions. When rotary valves are fully opened, air flows into the test matrix at a prescribed mass flow rate. The rotary valves were driven at the speed of 30∼240 rpm, and air temperature variation at both ends of the test matrix was measured. From measured data, the rate of heat regeneration of each test matrix was calculated.

抄録全体を表示

The adaptability of various materials for the matrix of a Stirling engine regenerator was evaluated inclusively, considering not only their heat transfer properties which were measured and reported in the 1st report, but also their flow resistances and void volumes. Also, the assumption which was applied in the 1st report to obtain the rate of heat regeneration from the measured temperature of air was examined by numerical analysis using the finite difference method and it was found that the assumption is adequate when the rotary valve speed is more than or equal to 240 rpm. To examine the accuracy of the experiment, the heat transfer coefficients which were obtained from the air temperature were compared to those calculated by generally known empirical formulae.

抄録全体を表示