To find a useful method for monitoring cavitation erosion arising in turbomachinery, several methods have been compared. There are mainly four typical methods : pure Al plate, soft Al foil, pressure-sensitive film and painting methods. However, the mutual relationship among them is yet unsettled. In this paper, therefore, the erosive area, where the erosion mainly takes place, was precisely detected by means of the four methods under the same flow conditions in typical separated flows around a slot. It was found that a large difference between the erosive areas is recognized, especially from the threshold levels, since each method inevitably has such an inherent characteristic threshold level against cavitation attacks.

View full abstract

This paper describes the deformation and collapse of a liquid column by shock wave loading and shock wave focusing. Experiments were conducted using a pressure-type 60mm×150mm cross-sectional shock tube equipped with a holographic interferometer. Shock wave focusing was performed by a reflected shock wave from a circular concave wall. A water column was placed at the focal point of the circular concave wall where the highest pressure was generated. Shattering processes of the water column were observed by means of holographic interferometry. Due to the high pressure of shock wave focusing, the shattering process in the earlier stage was promoted. The deformation of a water column was also accelerated. It is revealed that the surface wave over the water column contributes to the disintegration of the water column.

View full abstract

The Reynolds stress turbulence model has been investigated for more than 15 years. However, it has not been used widely in comparison with the k-ε model. One reason is that Reynolds equations are unstable because of the Eulerian characteristics and thus it is necessary to introduce a stabilizing procedure in solving the equations. I studied the applicability of the TVD scheme, which has been used successfully in compressible flow calculations, for incompressible Reynolds stress turbulence model calculations. A zero pressure-gradient turbulent boundary layer on a flat plate was calculated using the TVD scheme. It was found that the TVD scheme contributes to the stabilization of the calculation, especially in the case where the spatial resolution is not sufficient.

View full abstract

A recent method "wavelet analysis" is applied as a tool for analyzing time sequential quantities derived from chaotic systems. As typical examples, the present method is applied to two well known systems, the Lorenz model and the Rossler model. The main procedure consists of calculating the local cross correlations between the signal to be analyzed and a predefined waveform called a "wavelet" at various time scales. This makes it possible to unfold the signal into both time and frequency, whereas in Fourier analysis, results are obtained in the frequency domain. The present paper mainly focuses on understanding the properties of this method, and shows a way to interpret the results.

View full abstract

A method of simulation of a gas-solid two-phase flow of intermediate concentrations is developed. Particle diffusion due to inter-particle collision is treated by using the Direct Simulation Monte Carlo method. Four existing schemes of the DSMC method are compared and it is concluded that the Belotserkovskiy-Yanitskiy (B-Y) scheme is suitable for treatment of inter-particle collision in solid paricle systems. Particle collision with the solid wall is simulated by a semiempirical collision model. The gas flow is calculated by solving the compressible Navier-Stokes equation using the third-order accuracy upwind QUICK scheme and Reynolds stress turbulence model. This model is applied for "saltation" state particulate flow in a bend of a three-dimensional rectangular duct, and the calculation results are shown.

View full abstract

The characteristics of particle velocity in the main stream direction of solid-liquid two-phase flows in a swaying pipe are considered. The equation of motion for solid particles in a swaying pipe with consideration of virtual mass, the Basset term and frictional force between particles and pipe wall is derived. In addition, particle behavior is measured by means of digital image processing. Fluctuations of the velocity are clarified experimentally. The results of numerical solution of the above-mentioned theoretical equation correspond to experimental results.

View full abstract

A numerical analysis is conducted on heat transfer and fluid flow of a high-temperature jet including particles in plasma spraying, which is utilized to produce heatresistant, anticorrosion and abrasion-resistant coatings. Finite heat conduction analysis is adopted to consider the internal temperature distribution in a particle and further, the forces induced by the relative motion are added to the Stokes drag term in the equation of particle motion. The comparison between Lagrangian and Eulerian approaches to the evolution of velocity and temperature of alumina particles is presented. The equation of particle motion is calculated using the 4th order Runge-Kutta method, as is the residence time of particles in each calculation cell of the flow field using the bisectional method which was developed by us. Particles of three different materials (Al₂O₃, Ni, W) are considered and the coupling between two phases is examined on the basis of the momentum and energy exchange.

View full abstract

A prediction-based equation for heat flux through a multilayer insulator was derived from comparison of experimental results between room temperature and liquid nitrogen temperature. The employed multilayer insulator was a laminated material with a polyester net inserted between aluminized Mylar films. The prediction equation consists of one thermal radiation and two thermal conduction terms. The first conduction term is that of ordinary thermal contact conductance. The second conduction term depends on the self-compression of the multilayer insulation. The predicted values resulting from the obtained equation coincided fairly well with measured values.

View full abstract

Based on the detailed measurements of the unsteady boundary layer on the flat plate, which are affected by wakes generated from the moving circular cylinders, time-averaged characteristics of the boundary layer are calculated. Such characteristics are examined from viewpoint of the process of unsteady transition of the boundary layer. Furthermore, as for the time-averaged energy dissipation thickness of the boundary layer, some consideration is made to clarify the relation between the incident wakes and undesirable increase in total pressure loss, which is one of the important factors for aerodynamic designers of turbomachines.

View full abstract

It is well known that the Goertler vortex arises when a two-dimensional jet flows on the concave surface of y_0.5/R≥0.17∼0.27 (y_0.5 : half width of jet, R : radius of curvature). Much research has been done to clarify the condition which causes the formation of the Goertler vortex, the profile and the stability of the vortex, and so on. We know of no study to date on the three-dimensional, round wall jet along a concave surface. This study examines the equation of velocity profiles and clarifies the flow characteristics, distributions of mean and fluctuating velocities, decay of maximum velocity, spreads of jet, pressure distribution, and others, of the three-dimensional round-wall jet along a concave surface.

View full abstract

In the previous paper, the natural convections of thermo-electrically conducting fluids in a square cavity under a constant magnetic field are calculated using GSMAC-FEM in conjunction with the so-called B method. This scheme satisfies conservation laws of both mass and magnetic flux efficiently. In the present paper, we modify the conventional B method to raise the efficiency of calculation, while retaining the accuracy of calculation. Hydromagnetic cross helicity is introduced as the new physical quantity which accounts for the effect of interactions between the magnetic field and the velocity field. Using hydromagnetic cross helicity, we evaluate the difference in performance between the conventional and modified B methods quantitatively.

View full abstract

The equation of motion for a cluster of solid spheres moving in fluid at low Reynolds numbers is developed under the consideration of hydrodynamic interaction among the spheres. The hydrodynamic forces of the spheres in the equation of motion are estimated using the relative velocities. The relative velocities are derived by applying the usual reflection method for the Stokes flow field to the newly developed reflection method for the Oseen flow field. In order to discuss the effectiveness of the above-described method, experiments for two equal spheres falling freely in quiescent glycerol are carried out. The width of the test tank with a square cross section is 100 times the sphere diameter. The positions of the spheres for time lapse are measured using a stroboscope-camera system. The numerical results show good agreement with the experimental results.

View full abstract

This paper presents a comparative study of several numerical methods which have been proposed for solving two-dimensional incompressible viscous flows. These methods are the vorticity stream function method, the MAC method on a staggered grid, Chorin's artificial compressibility method with/without convergence acceleration and Abdallah's method on a non-staggered grid. The steady laminar flow in a square driven cavity and a two-dimensional channel with 90° bend are chosen as test problems. It is found that all of the methods give practically the same computed results except for the pressure distributions in the two-dimensional channel problem obtained by the vorticity stream function method. The numerical experiments show that the artificial compressibility method with convergence acceleration has the best convergence rate.

View full abstract

An experimental study was made to clarify the fundamental characteristics of the tube side flow rate distributions in a multitube heat exchanger with several tube passes. In this experiment, 2 and 4 tube passes divided a tube bundle consisting of a hundred branch tubes into two and four parts, respectively. Experimental parameters considered were the mean flow rate of the tube bundle, the inlet and outlet nozzle direction, the nozzle diameter, and the length of the header without nozzle. Results show that the flow rate distributions of the 1st pass changed markedly with the inlet nozzle direction and the nozzle diameter, while those of the 2nd pass were almost the same for the various flow conditions and became more uniform. Also, a suitable length of the header without nozzle existed for the flow rate distribution of the tube bundle.

View full abstract

The double structure of cells is a very flexible, powerful tool to find the cell in which a particular molecule is located using the Monte Carlo Direct Simulation. The flow field is first divided into very small rectangular grid cells (small cells). Since the macrocell in which intermolecular collisions are counted and macroscopic properties are sampled is constructed of many grid cells, the simulated region of physical space can be divided freely into a network of macrocells. Although a table indicating the macrocell to which each grid cell belongs is essential in the double-cell structure, it is troublesome work to obtain a reasonable table, especially for complicated flow fields. A computer program for assembling grid cells and constructing a network of macrocells is generated by both a common computational method and a method utilizing a neural net. Cell networks calculated with these methods are shown and discussed.

View full abstract

A numerical technique for the computation of three-dimensional turbulent flows through a turbine stage with bucket tip clearance is presented. To calculate nozzle and bucket flow regions simultaneously, steady interaction is assumed on the connecting boundaries. To consider the exact flow-region geometry of the bucket tip clearance, an H-type computational grid system, which is also used in the nozzle and bucket flow regions, is used in the tip clearance region. A finite volume method is used to obtain the spatially discretized governing equations, while the damping surface technique is employed for the time integration. In the present analysis, a two-equation model of turbulence is introduced to estimate the turbulence effect. In order to assure the effectiveness of the present method, computations are carried out for the flow through turbine stages of different tip clearance heights. The results clearly show secondary flow phenomena such as the tip leakage vortex and passage vortex motions in the bucket flow region. The effects of the clearance height on the turbine stage flow fields are also well predicted qualitatively by the present method.

View full abstract

This paper presents the three-dimensional finite element formulation for calculating aerodynamic forces acting on oscillating linear cascade blades in a subsonic flow. Under the assumption of the potential flow, the velocity potentials at nodes are determined as the solution for the simultaneous algebraic equations. The expressions of acoustic and wake potentials are also given and used to determine upstream and downstream boundary conditions. The present formulation is applied successfully to a couple of three-dimensional geometries in the accompanying paper.

View full abstract

In the present paper, the three-dimensional finite element method developed to calculate the aerodynamic forces acting on the vibrating cascade blades in a subsonic flow is applied to various flow conditions and geometrical configurations. The present FEM formulation using pentahedra instead of tetrahedra reproduces perfectly two-dimensional solutions when applied to the spanwise uniform model. Calculations have been made for three-dimensional configurations with various dihedral angles, sweep angles and taper ratios. In most cases, three-dimensional effects due to taper on steady and unsteady aerodynamic forces appear as the effects of decreasing spanwise variation of steady and unsteady forces, respectively, when the angle of attack is uniform along the span. Three-dimensional effect due to taper becomes small when the angle of attack is nonuniform along the span. The effect of dihedral is small compared with the effect of sweep or taper. The effect of sweep on unsteady aerodynamic forces becomes extremely conspicuous when the steady loading is nonuniform along the span.

View full abstract

To clarify the running characteristics in oscillating flow of a cross-flow turbine with two axisymmetric nozzles, experiments are carried out using a turbine test equipment in which wavelike up-and-down oscillating flow conditions are simulated. On the other hand, the dynamic characteristics of this turbine are calculated numerically on the basis of experimental data obtained from experiments in steady one-way flow conditions. The experimental results are compared with the numerical results and good agreement between them is demonstrated. Furthermore, the effects of load control and period of oscillating flow on the running characteristics are investigated.

View full abstract

Two different types of air bearing pads have recently been used in the most precise machine : feeder hole-type and groove-type, which can be externally pressurized. Practically speaking, groove-type pads are more common. However there are few theoretical analyses or experimental studies on air thrust bearings. Therefore, in the present paper we describe the measurements of the surface pressure on the rectangular groove-type pad and the static characteristics of air-lubricated thrust bearings, and discuss the stiffness of air bearings. Furthermore, the dependence of the pressure distribution on the stiffness with the variation of the effective area, the effective area ratio, and the groove depth is discussed.

View full abstract

Ultrasonic enhancement of heat transfer on a narrow surface was measured by changing the width of the surface from 8 to 0.1 mm. Ultrasonic power of 600 W with a frequency of 40 kHz was used. Heat transfer on the narrow surface without ultrasonic vibration was correlated by means of the experimental equation for thin wire. The cavitation intensity was measured by means of the cavitation erosion loss of aluminum foil of 15μm thickness. The effects of acoustic streaming and cavitation were separated by this measurement. Heat transfer by acoustic streaming was predicted through the forced convection. Enhancement by cavitation was explained by the turbulence heat conductivity of the microjets.

View full abstract

This paper describes the comparison of the flow structure over the rows of two-dimensional ribs of various rib shapes on the ground plane. The pitch between the centers of two adjoining ribs was varied for several rib shapes. Six two-dimensional rib shapes were used in this experiment. The time-mean velocity and turbulence intensity were measured by laser Doppler velocimetry. The auto-correlation was obtained using a data processing system and Fast Fourier Transform analyzer connected to a hot-wire anemometer. The mean temperature was measured with thermocouples and the local heat transfer coefficient was obtained from these data. The turbulence intensity and heat transfer coefficient become largest and the pressure loss is minimum for the semi-elliptic ribs, among all the rib shapes used in this experiment. As a result it is concluded that semi-elliptic ribs aligned at pitch ratio S/D=7 have the optimum shape to promote heat transfer.

View full abstract

Turbulent quantities were measured in a fully developed pulsating pipe flow heated with a constant wall heat flux using a hot-wire/cold-wire technique. The pulsating frequency was changed to cover the natural bursting frequency, while the relative pulsating amplitude was kept almost constant at 12%. For comparison, the measurement was also repeated for fully developed steady pipe flows at the same Reynolds numbers. It was found that the time-averaged characteristics of the turbulent flow and thermal fields near the wall are apparently insensitive to the imposed pulsation.

View full abstract

The surface temperatures of the inner lamp and the outer glove of a spherical lighting fitting, the surfaces of which are painted black, are measured. From the results the average cnonvective heat transfer coefficients between the inner lamp and the outer glove and on the outer surface of the glove are obtained. These data are correlated with the aid of existing equations for two concentric spheres and the outer surface of a single sphere. The relationships between the maximum and mean temperature on the lamp and the glove are also obtained. By the use of these equations a method for the optimal thermal design of spherical lighting fittings is proposed.

View full abstract

Increasing attention has been focused on the development of small-scale boilers. One approach is to adopt the concept of the tube-nested combustor developed by Ishigai et al. Experimental investigation on heat transfer has been conducted using such a simulated tube-nested combustor with emphasis on the spatial distribution of the heat transfer coefficient in the tube banks. Heat transfer coefficients at the impinging region of the jet are significantly higher than those in the uniform flow. This heat transfer coefficient decreases with the increase in the distance from the inlet, and approaches the value obtained under uniform flow. The transverse distribution depends significantly on the flow pattern of the jet in the banks.

View full abstract

The enhancement of longitudinal heat transfer by means of fluid pulsation in a pipe has been investigated analytically and numerically, including the transient state. The effects of pulsation amplitude, frequency and pipe length on thermal properties such as effective thermal diffusivity and delay time are clarified. Their effects on numerical calculations are also presented, and suggestions for efficient numerical calculations of this problem are made concerning the combination of parameters.

View full abstract

A molecuar dynamics simulation has been carried out on the melting behavior of an atomic layer irradiated by thermal radiation. One thousand atoms in this layer are connected to each other by the network of a two-body interatomic potential function. The irradiated electromagnetic radiation is characterized by the wavelength and the phase as well as by the amplitude. The radiation energy is absorbed volumetrically in the atomic layer, and transformed to kinetic energy of the atoms. A solid-liquid interface is found to appear at the surface of the layer, and to propagate into the inner part of the layer. Clear difference in atomic configuration and atomic movement has been observed between the solid and liquid phases. Successive irradiation causes the evaporation of atoms in the liquid phase, and atoms in the liquid and gas phases reach their critical state in the closed system of this simulation model.

View full abstract

This study focuses on the second-moment closure of turbulent scalar transport. Special attention was given to the characteristic time scales of scalar-pressure gradient correlation and dissipation terms in the turbulent scalar flux transport equation. The correlation coefficients of those terms and the turbulent scalar flux were introduced in order to represent influences of the Prandtl number (Pγ=v/α), time scale ratio (R=k_θε/ε_θk) and turbulent Reynolds number (Re_t=k²/vε). An effort was also made to take into account effects of the turbulence anisotropy produced by a strong mean shear rate. The proposed model provides excellent predictions of scalar flux budgets in isotropic turbulence, homogeneous shear flows and fully developed channel flows.

View full abstract

The present paper deals with size effects upon burnout heat flux on a heated ribbon. Saturated water was boiled on an upfacing heated ribbon under atmospheric pressure. The length and the width of the ribbon were varied over a wide range, and burnout heat flux variation with the ribbon width was studied experimentally. The departure diameter and the hovering period of a vapor mass, which may both be closely related to burnout heat flux, were measured using a video tape recorder and an electric resistance probe. The experimental results show that with decreasing ribbon width, burnout heat flux increases while the departure diameter, as well as the hovering period, decreases. The theoretical analysis developed for the motion of a vapor mass explains well the experimental results of departure diameter as well as those of hovering period. However, the variation of burnout heat flux cannot be explained by only the variation of hovering period. Two possible mechanisms of burnout are proposed and discussed based on the suppression of macrolayer formation and the microlayer contribution to heat transfer. A different character of burnout heat flux on heated wires is also discussed.

View full abstract

We have simulated the heat transfer in aquifer thermal energy storage by a full-scale code that has reflected local properities in the underground. Although this simulation code is very useful in understanding heat transfer accurately, it is not economical under uniform properties. In this paper, the full-scale code is converted to enable use of the complex potential of its flow as a nodal system in order to analyze Lagrangian coordinate systems. This results in quite rapid and accurate simulation and enables us to run it on a personal computer. Moreover, this code is best suited for consideration of the anisotropy of thermal dispersion due to the flow domain.

View full abstract

In many countries, especially in Europe, the study of heat transfer in aquifer thermal energy storage (ATES) has been carried out in order to evaluate the utility of natural energy for the future. Simulation has proven useful for clarifying mechanisms of heat transfer in underground water, and much progress in this area has been made to date. However, the available simulation methods cannot be applied to the analysis of heterogeneous or anisotropic subjects in nature. For this reason, these codes do not enable analysis of ATES in practical terms. Consequently, we have developed two simulation codes. One, called the "full-scale code", can be applied to any case. The other, called the "rapid code", enables rapid analysis of ATES while incorporating the consideration of heterogeneity in thermal dispersion due to the flow domain. In this paper, we present the full-scale code ; the rapid code will be addressed in the following paper.

View full abstract

We studied premixed gas combustion experimentally using annular nozzles which have coaxial bluffbody in a pipe. Cross-sectional area of nozzles is kept constant, while the opening width between the inner surface of nozzles and the outer surface of bluffbody is varied. We find that an annular nozzle has four flame patterns : flame attached to outer edge, flame attached to both outer and inner edges, flame attached to inner edge and lifted flame. Each flame pattern is dependent on mixture gas velocity and the equivalence ratio. Furthermore, each flame pattern region is affected by the ratio of the inner diameter to the outer diameter of the annular nozzle. This ratio has an optimum value, which gives maximum velocity at the lean limit in the region of flame attached to inner edge.

View full abstract

Under atmospheric pressure, the tendency of members of unsaturated series to smoke generally decreases with the number of carbon atoms. The effect of ambient pressure on the smoke point has been studied to explore the tendency of alkene fuels to smoke at high pressures. A wick lamp used in this experiment is standardized by JIS. The smoke point was determined for several alkene fuels under various ambient pressures from 0.2 MPa to 1.35 MPa. It has been found that the dependence of ambient pressure on the smoke point of alkene fuels is almost the same as in the case of alkane fuels. That is, as the ambient pressure increases, the smoke point decreases sharply up to the pressure of 0.5 MPa, beyond which it decreases gradually. However the effect of ambient pressure on the smoke point of alkenes is not very strong compared with that of alkanes. In alkenes, the smoke point takes the maximum value with a certain number of carbon atoms at ambient pressures of 0.3 MPa and above. In light of this finding, it can be concluded that there may exist a number of carbon atoms at which the tendency to smoke is minimized in alkenes.

View full abstract

An experimental study was conducted to examine NO_x emission of a premixing-prevaporizing flame stabilized with a bluff body. The liquid fuels, such as kerosene and light oil, were sprayed into the premixing-prevaporizing chamber with an air-atomizing nozzle. The size distribution of fuel spray, vaporizing time and heterogeneity of the mixtures were varied by varying the degrees of atomization and the length of the premixing-prevaporizing chamber. Velocity, concentration and Sauter mean diameter of the fuel spray in the vaporizing process were measured by means of a particle dynamics analyzer (PDA). The NO_x emissions from premixed -prevaporized flames decreased with decreasing size of sprayed fuel droplets and increasing vaporizing rate. It was possible to reduce the NO_x emissions from kerosene flames to the same level as methane-air premixed flames. While light oil has the lowest vaporizing rate, the temperature profile of its flame became similar to that of the methane flame and its NO_x emission level also approached that of the methane flame with the promotion of vaporization.

View full abstract

The new technique for the simultaneous measurement of fuel vapor concentration and droplets density was adopted to the evaporating diesel spray injected into a high temperature and high pressure environment. This measuring technique was based on the principle of ultraviolet light absorption caused by the fuel vapor and visible light scattering caused by the droplets in the α-methylnaphthalene diesel spray. It was found that in the case of the free spray the distribution of the fuel vapor concentration is strongly affected by the ambient temperature, the specifications of the injection nozzle. In the case of the impinging spray, the optimum impingement distance exists to facilitate the droplets evaporation and fuel-air mixing on th impingement wall.

View full abstract

The characteristics of the ambient gas entrainment into an evaporating diesel spray were analyzed by using the results of the distributions of fuel vapor concentration and droplets density, which were measured by the new technique developed in this study. It was found that the characteristics of the ambient gas entrainment was strongly affected by the ambient gas temperature and impingement distance. The total mass of the ambient gas entrained into the evaporating spray impinging on a flat wall was grater than that of the free spray at the same time from when the injection started. In the case of the impinging spray, the optimum impinging distance exists for facilitating the entrainment of the ambient gas.

View full abstract