The velocity of a free-falling sphere in the acceleration range in water/fine-solid particles has been measured by means of four ultrasonic transducers. Carbon black suspensions in the concentration range Cw=0.3∼0.5wt%, and in SiO₂ suspensions of C_w=0.5∼1.0wt% were tested in this experiment. Six spheres between 7.93mm and 12.70mm were tested. From the comparison of the experimental results with the calculated results of the general equation of motion considering the Basset term, it was shown that the drag reduction occurs in carbon black suspensions.

View full abstract

Rarefied gas flows in various situations are calculated successfully by the direct simulation Monte Carlo method. In the simulation, the Maxwell model, where a gas molecule reflects diffusely with the probability a and reflects specularly with the probability 1-a, is widely used for boundary conditions on solid surfaces. However, the value of a is determined empirically and varies greatly with conditions such as degree of contamination and temperature of the surface. Rational prediction of the value and analysis of the interaction between gas and solid surface are required. In this paper, the behavior of a gas molecule which collides onto the solid surface is simulated by the molecular dynamics method. The numerical results reveal that the scattering behavior of the gas molecule is neither specular, diffuse, nor Maxwell-type reflection, and that the sticking probability is affected by the initial gas velocity and the potential well depth.

View full abstract

Rotating cavitation is analysed by means of an actuator disk method. Quasi-steady pressure performance, mass flow gain factor and cavitation compliance are taken into account. Three types of destabilizing modes are observed : rotating cavitation propagating faster than the impeller speed, rotating cavitation propagating opposite to the impeller rotational direction, and rotating stall propagating slower than the impeller speed. It is shown that both of the rotating cavitations are caused by the positive mass flow gain factor, while rotating stall is by the positive slope of the pressure performance. Stability and propagation velocity maps are given for the two types of rotating cavitations in the mass flow gain factor-cavitation compliance plane. The results of the analysis and the experimentally observed rotating cavitations are discussed.

View full abstract

The motion of a single bubble is simulated when the surrounding pressure decreases step-wisely. In this simulation, the conservation equations for mass, momentum and energy are solved directly in order to estimate the effect of internal phenomena on bubble motion. At the same time, the mist formation in the gas phase and the diffusions of heat and noncondensable gas in the liquid phase are taken into account. The numerical results for several cases reveal that nondimensional transport coefficients have large effects on the distributions of temperature, concentration of vapor and the mist formation. As the initial radius becomes smaller or depressurization ratio becomes larger, bubble motion is less influenced by mist formation due to the heat penetration through the bubble wall by heat conduction.

View full abstract

This paper describes the deformation and collapse of a liquid column due to a shock wave and shock-wave-induced high-speed air flow. Experiments were conducted using a pressure-type shock tube equipped with a holographic interferometer. A shattering process of the water column is observed using holographic interferometry. The shattering process is characterized by three stages : 1st stage, deformation of a water column ; 2nd stage, appearance of crests at the separation point of a boundary layer over the water column and onset of surface waves ; 3rd stage, breakup and atomization due to the surface wave. Time variations of the mass, and longitudinal and transverse diameters of a shattering water column are presented for incident shock wave numbers Ms=1.18, 1.47 and 1.73. The structure of a flow around the shattering water column is discussed in detail.

View full abstract

Pulsatile flow in a curved pipe enhances gas or substance transport because of the existence of secondary flows, and provides more mixing in the liquid phase than do steady flows. In this study, numerical analysis was performed on fully developed pulsating flow in a curved pipe under flow conditions similar to those characteristic of the blood flow in an aortic arch or in a booster lung. The calculations were performed over a wide range of Dean numbers De and Womersley numbers Wo for a fixed curvature ratio δ=0.05, and Schmidt number Sc=2900. The isodensity contour of the axial direction having a thin boundary layer near the wall was shown. Furthermore, it was revealed that the time- averaged Sherwood number (Sh)^^^- increased by about 10 times due to the convection effect of secondary flows in the case of high Dean number flows compared with low Dean number flows, and that in the case of pulsatile flows, the time- averaged Sherwood number approached the values that of steady flows when the volumetric flow rate β equaled 10.

View full abstract

In this study, we have considered a reverse transition phenomenon, i.e., the disappearance of velocity fluctuation in the laminar-turbulent transition of spherical Couette flow with only the inner sphere rotating. This phenomenon occurs within a limited clearance ratio region around 0.14, where the rotating disk type's disturbances, such as Stuart vortex, ring vortex and shear waves, can appear simultaneously with the rotating cylinder type's disturbances such as spiral Taylor-Gortler vortex and travelling waves. When the reverse transition occurs with an increase of Reynolds number, the damping of fluctuating velocities begins in the order of u, v and w, where u, v and w are the fluctuating velocity components of azimuthal, radial and meridian direction, respectively. The R. M. S. values of v and w are larger than that of u for the case of travelling waves.

View full abstract

We have experimentally investigated the influence of the shape near the trailing edge of a flat-plate blade on the periodic velocity fluctuation in the near wake of the single rotating blade. The results show that the characteristics of the near wake in the rotating flow field are similar to those in the uniform flow field; i. e., the periodic velocity fluctuation did not occur when the suction surface of the blade was scraped near the trailing edge, and did occur if the pressure surface was scraped. In the latter case, the two types of trailing edge separations occur. The flow always separates at the trailing edge of the blade on the suction side, while the flow separation alternately originates from two different points, i. e., from either the starting point of the scraping or from a fixed point on the scraped surface.

View full abstract

The behavior of the excitation caused by longitudinal vortices shedding from two cruciform circular cylinders with an equal diameter d is investigated using a wind tunnel. When the gap s between two cylinders is less than d/2, longitudinal vortices shed periodically and a fluctuating lift force is exerted on the upstream cylinder. The nondimensional frequency, St_M, of the longitudinal vortices is independent of the Reynolds number. The value of St_M is 0.06<St_M<0.09 when s/d<0.25 (region A) and St_M≒0.04 when 0.25<s/d<0.5 (region B). The fluctuating lift force on the upstream cylinder is estimated from the pressure on the cylinder surface using a phase-averaging technique. The result shows that the flow near the crossing of the cylinder is dominated by four trailing vortices in the region A and by two necklace vortices in the region B. These vortices shed an anti-phase on the opposite side of the upstream cylinder, causing an alternative lift force to be exerted on it. The abrupt change of St_M at s/d=0.25 is caused by the fact that these two types of vortices interchange their role to excite the upstream cylinder.

View full abstract

In previous research, a simple finite-element method using simultaneous relaxation of velocity and the Bernoulli function was presented to simulate three-dimensional flow prblems. In the present paper, this scheme is expanded to consider heat convection problems. In order to reduce the computational storage, coefficient matrices are calculated by simple integration formulas for an eightnode isoparametric element. Natural convection in horizontal concentric annuli is calculated to certify the scheme. The present method is stable in heat convection analysis and the numerical results agree well with the experimental results.

View full abstract

A numerical scheme for analyzing the steady two-dimensional incompressible viscous flow using a general curvilinear coordinate grid is proposed. In this scheme, the unsteady Navier-Stokes equations are solved by a convective-difference scheme using a staggered square grid in transformed space. An elliptic equation of pressure is solved by the Tchebyscheff SLOR method. The substantial derivative term in the convective-difference scheme is integrated along a path line and the values at the upstream end are interpolated considering a TVD concept. As numerical examples, the backward-facing step duct and U curved duct flows were calculated. The calculated results show that the scheme has good accuracy as a second-order scheme and is speedy in reaching for the steady condition.

View full abstract

This paper reports the development of a sheet carrying mechanism by means of the progressive wavy motion of a thin elastic membrane which is sustained by magnetic fluid. The sheet carrier consists of a magnetic fluid layer in a shallow rectangular vessel, a thin elastic membrane covering the magnetic fluid, and digital electronic equipment to drive and control a series of magnetic coils. By applying the magnetic field in the form of sinusoidal waves, progressive wavy motions of the magnetic fluid and the elastic membrane are excited, and the sheets of paper or plastic film can be carried on the surface of the membrane. An interesting transportation behavior and its dependence on the wavelength and the frequency of the progessive waves are examined experimentally.

View full abstract

A Lagrangian-type numerical simulation was attempted for a two-dimensional fluidized bed. The availabilty of computers with large capacity opens new possibilities for treating individual particles in a fluidized bed. The solid phase in the bed was obtained from calculations of the motion of individual particles using the ordinary Newtonian equations, contrary to most previous simulations where the solid phase was regarded as a continuum. The gas phase was solved using the Eulerian equations for local variables, taking into account the gas-particle interaction. The particle-particle interaction through contact forces is a dominant factor in the case where the particle concentration becomes larger, such as in a fluidized bed. The contact forces are modeled by mechanical units such as a spring, dashpot and friction slider, in the present work.

View full abstract

Fuel gas injection is used for gas diesel engines. It is very important to investigate the behaviour of the unsteady jet. In this study, the concentration field of a suddenly started cold methane gas jet into still air was nonintrusively measured by the infrared absorption method with a 3.39μm He-Ne laser. Furthermome, the equations of mass, momentum, energy, species (methane and air) and the k-ε turbulence model were solved numerically. The concentration field was obtained and compared with the experimental results. As a result, it was found that the global trends of the jet, that is to say, penetration and averaged concentration, are predicted well, though the unsteady concentration near the jet axis is not predicted because the turbulence model is based on the time-averaged one.

View full abstract

The switching mechanisms in devices such as wall attachment fluid amplifiers, turbulence amplifiers and beam deflection amplifiers have excellent properties, and many investigations have been reported. In turbulent opposed jets, it is found that the impact position can be controlled by the turbulence in the jet center. In this paper, the performance of a new fluidic amplifier applied to the above mentioned phenomenon is described. The new fluidic amplifier consists of an outer pipe, two opposed nozzles, a small nozzle to generate control flows and a separator. The impact position generated by the opposed jets is controlled by the small control flow by way of the small nozzle. The new fluidic amplifier can be driven by small control flow, and as a result, is a useful component in a complex circuit.

View full abstract

Coriolis effects on fully developed turbulent flow in a two-dimensional channel rotating about an axis perpendicular to its axis are considered. The Coriolis force has stabilizing/destabilizing effects on turbulence, and the mean velocity distribution changes accordingly. Experimental and numerical studies on the velocity characteristics have already been conducted by other researchers for various conditions. However, we cannot assemble the overall picture of the Coriolis effect on the velocity distributions. In this paper, analytical studies have been performed using dimensional analysis, which indicates that the three lenght scales and their ratios constitute the dominant parameters governing the flow. The flow regime space has been shown to account for the overall Coriolis effect on the flow. Furthermore the velocity laws for various flow regions have been derived.

View full abstract

In order to clarify the turbulence behavior in impeller channels of turbomachinery, effects of the Coriolis and centrifugal forces on the turbulent boundary layer on the concave and convex surfaces were studied experimentally, using a rotating channel of rectangular cross section whose aspect ratio was 2 : 1. Velocity profiles and turbulence intensities were also compared with those for a channel having an aspect ratio of 4 : 1. In the present rotating channel it was found that secondary flows generated by the centrifugal and Coriolis forces have a large effect on the time-averaged and turbulence components in the boundary layer along the concave and convex surfaces, and that the suppression of turbulence expected on the lower-pressure side of a two-dimensional flow is not observed because of the accmulation of lower energy fluids and the interference of the fluids with the primary flow.

View full abstract

The general hot-wire response equations of the triple-sensor hot-wire probes were introduced. Assuming normal distributions for the turbulence velocities, numerical experiments were carried out on the measurements of the circular free jet and the weak isotropic turbulent flow. The results are summarized as follows : the measured mean velocity and Reynolds stresses are exact when the turbulence intensity is weak, as far as the values of the yaw and pitch factors are known ; the magnitudes of the relative errors of (u²)^^^- and (uv)^^^- by the triple-sensor probes are nearly equal to those by the X-wire probe when the turbulence intensity is strong ; and the maximum value of the effective cooling velocity of a turbulent flow in which the turbulence velocity always stays inside the acceptable domain is approximately 1.4 times the value of the mean velocity.

View full abstract

The main purpose of this study is to verify experimentally the method proposed by the authors for estimating the position of a detected leak by applying a fluid transient model to a real-time pipeline leak detection system. The method considers the flow rate and pressure profiles changing along a pipeline in a transient manner. Experiments were carried out on a pipeline of 115m length and 3/4 inch diameter using dry air. Pressure and Reynolds number examined are from 0.098 to 0.127MPa and from 3000 to 12000, respectively. The result shows that, in both steady and transient conditions, the proposed method gives a more accurate estimation of a leak position than do conventional methods which do not consider flow rate or pressure profiles along a pipeline. Additionally, the influence of pipe roughness on the estimation is discussed.

View full abstract

Converging shock waves in solids are numerically simulated using the random choice method (RCM), which has been developed for gases. Initially, a Riemann solver for fluidlike solids with the Gruneisen-type equation of state is constructed. It is incorporated into the RCM, and is applied to the cylindrical shock tube problem in solid copper with a driving pressure of 20 GPa. Spacial distributions of pressure, density and particle velocity show that the steepness at the shock front is maintained both in the converging and reflecting stages. Numerical results are compared with those of the finite difference method (FDM), showing the superiority of RCM over FDM. It is shown that the pressure on the shock front and the total energy contained in the central circular area in the reflecting stage are much larger than those in the focusing stage.

View full abstract

The SOLA (solution algorithm) method is one of the most widely used methods for analyzing incompressible viscous flows. However, this method often requires a number of iterations to obtain a converged solution for a practical problem on transient flow. Hence, a theoretical analysis of this method was carried out in the present study to clarify the reason why this method can give converged solutions. It was proved that the SOLA method is identical to the SMAC (simplified marker and cell) method in the interior cells of the computational domain. On the other hand, it is not identical to the SMAC at the cells adjacent to the boundary cells. Hence, a modified SOLA method which improves the computational efficiency was developed by taking into account boundary conditions in the iteration process. It was demonstrated that the modified SOLA could analyze transient flow more efficiently than the original SOLA.

View full abstract

Computational time for numerical simulation of transient incompressible viscous flow is governed by numerical methods to solve the Poisson equation of the pressure. In order to improve the computational efficiency of the SOR method, a mathematical analysis was conducted in the present study to derive the spectral radius, i. e., the maximum modulus of the eigenvalues of the iteration matrix for a discrete Poisson equation in staggered mesh system. The derived spectral radii for the point, line and area Jacobi iteration matrices are applicable to the Dirichlet, the mixed boundaryvalue and the Neumann problems in an n-dimensional rectangular region. The validity of the optimum relaxation parameter of the SOR method, which was calculated by the derived spectral radii, was confirmed by numerical experiments. Furthermore, a simple estimation method of the optimum relaxation parameter for the non-rectangular region was presented for transient flow analyses.

View full abstract

Characteristics of a water sprinkler which sprays rectangularly distributed water on railway tracks on which high-speed trains run, and keeps the snow in wet condition in order to protect the running gear and underfloor equipment of vehicles from the snow damage are discussed. Field tests show excellent results of more than 70% in the equality coefficient and more than 60% in the sprinkling efficiency. The new sprinkler covers the entire track area instead of a limited circle covered by the conventional sprinkler so that it will be possible to operate the Tokaido Shinkansen trains in winter at a faster speed than at present.

View full abstract

The effects of system pressure in gas-liquid two-phase flow cause a significant change in gas-liquid interfacial structures. The purpose of this study is to investigate the pressure effects on the void fraction from bubble flow to the annular flow region. The void fractions were measured using a needle probe method. It was found that there is a region in which the gradients of the mean void fraction vs superficial gas velocity curves show a minimum point or a changing point near the disappearance point of the liquid slug under constant superficial liquid velocity. In this study, system pressure in the closed-loop system was changed substantially from 0.3 to 20MPa at constant fluid temperature in vertical upward flow.

View full abstract

An experimental study is reported on convective boiling heat transfer for nonazeotropic refrigerant mixtures of HCFC22/CFC114 inside a horizontal smooth tube. The test evaporator is of the double-tube type in which the heating water flows in the outer annulus, and the local heat transfer coefficients are measured for both counter and parallel flow. Based on the equation, which correlates the heat transfer coefficient for pure refrigerants as the sum of forced convective contribution a_cv and nucleate boiling contribution a_nb, an empirical correlation equation for the mixture is developed with the data corresponding to the annular flow pattern. The mean deviation between the calculated and measured heat transfer coefficients is 8.9% for the present experimental data and 7.1% for data obtained by Jung. The equation reveals that the contribution of the nucleate boiling to the total heat flux is much smaller than that in pure refrigerants, while the convective heat transfer coefficient is almost the same.

View full abstract

An experimental study has been conducted to investigate the heat transfer of evaporation of a water droplet on a hot stainless-steel cylinder coated with a refractory paint. In order to measure temperature changes of the cylinder, four thermocouples are used. Their radial positions are 0, 1.5, 3.0 and 4.5mm away from its central axis, respectively. The distributions of temperature and heat flux of the heat transfer surface are obtained with the inverse calculation of two-dimensional heat conduction in both coating and metal regions. In the calculation the liquid-solid (L-S) contact surface is assumed to be a circle whose center is located at the cylinder axis. The measured temperature drops rapidly upon L-S contact, and after attaining a minimum, it increases slightly and then remains constant for a while. In the L-S contact region, the surface temperature distribution is almost uniform, while the heat flux is higher in the boundary than in the center. The total heat transferred through the heat transfer surface agrees well with that required for the evaporation of a droplet.

View full abstract

An experiment on steady transition boiling of water was performed under atmospheric pressure on a horizontal copper disc 10mm in diameter. An impedance probe was used to study vapor-liquid behaviour on the heated surface. It was found that there are two different heat transfer modes in transition boiling, depending on the wall superheat. The macrolayer evaporation model, which was first proposed by Katto and Yokoya, is verified experimentally based on the present authors' heat flux data in the liquid occupied region of the heated surface. The transition boiling heat transfer at low wall superheat can be well explained by the model.

View full abstract

This study has been conducted to pursue heat transfer enhancement in a convective field by utilizing gas-solid suspensions controlled by a steady electric field. The gas-solid suspensions consisting of air and 40μm glass beads flowed vertically upward in a channel whose walls were subjected to high voltage potential. The observations of particle trajectories and the measurements on heat transfer rate were systematically undertaken. Firstly, the primary behaviors of charged particles under the action of the Coulomb force in an electrohydrodynamic (EHD) field were experimentally clarified. Secondly, the heat transfer rates from both surfaces of the channel wall were augmented with increasing applied voltage and loading ratio, and the mechanism of enhancement was found to correspond positively to the number density of the particles colliding with the wall surface. On the basis of the results, the availability of utilizing electric field was made clear in comparison with a single-phase flow under the constant rate of power consumption.

View full abstract

The porous insulation layer may include inhomogeneous high-porosity regions near the bounding walls, which introduce additional thermal resistance and enthalpy transfer, i.e. channelling effects. The present paper deals with the channelling effects by considering effective fluid layers as models for the high-porosity regions. The paper is divided into two parts. First, boundary conditions at the interface between the porous and fluid layer are discussed, and the Beavers-Joseph condition is extended to take account of rapid velocity change in the porous layer near the interface. Second, on the basis of the extended boundary condition, an analysis is made of the effects of both buoyant channelling flow and the near-wall thermal resistance on the heat transfer characteristics. It is shown that the heat transfer is reduced by the additional thermal resistance, while it can be enhanced by the enthalpy transport due to the channelling flow. Discussion is also given of the critical thickness of the effective fluid layer to eliminate the channelling enthalpy flow.

View full abstract

Heat transfer from small heating elements on a substrate has been studied experimentally and analytically. In one of the experiments, the heating elements are diodes in an actual IC chip. In another set of experiments, heaters are small thin films deposited on a glass plate. In both experiments, the substrates are cooled by an impinging jet of air or channel flow. The experimental data indicate that heat conduction from the heating element to the substrate is an important factor in the determination of the maximum temperature of the element, while the surface heat transfer determines the bulk temperature of the chip. The numerical analysis clarifies that a chip surface with an area much larger than the element size is necessary for the heat transfer to the air flow.

View full abstract

An experimental study was conducted to determine if the deterioration in heat transfer caused by the insertion of a large eddy break-up manipulator (LEBU plate) could be recovered by attaching a half-delta wing vortex generator to the LEBU plate. The vortex generator was found to work well to augment the wall heat transfer over a large streamwise distance, although the LEBU plate itself is expected to eliminate the large-scale eddy motion in the boundary layer. Some geometric parameters of the inserted LEBU plate attached to the vortex generator were varied in several steps. Within the experimental ranges of such parameters, the height and angle of attack of the vortex generator were found to seriously affect the magnitude of heat transfer augmentation, which becomes more pronounced as both of their values are increased.

View full abstract

Experimental investigations are made on the enhancement of forced convection heat transfer using thin inclined plates in a rectangular channel. The channel constructed with these plates has diverging and converging parts alternating along the flow passages. Heat transfer coefficient and pressure drop are measured for air flow at Reynolds numbers ranging from 1.2×10⁴ to 4.8×10⁴. Performance comparison is made under the condition of constant pumping power. Large enhancement in the heat transfer coefficient is achieved as the angle of inclination increases, with accompanying large pressure drops. The channels with clearance at the peaks of converging-diverging parts are tested in order to decrease the large pressure drops. When the clearance is perpendicular to the mainstream direction, an improvement in performance is observed.

View full abstract

A two-equation heat-transfer model for high-Prandtl-number fluid flows has been developed from the scope of the original (t²)^^^- ε_t model of Nagano and Kim (1988). The proposed model has been appraised in comparison mainly with the existing experiments on water (Pr=5.9), aqueous ethylene glycol (Pr=14.3) and oil (Pr=95). It is shown that the present model predicts quite well the experimental behavior of fundamental quantities in turbulent heat transfer, i.e., the Nusselt or Sherwood number, temperature profiles and intensity of temperature fluctuations. Especially, the prediction of the r.m.s. intensity of temperature fluctuations is in good agreement with the most recent and reliable experimental data of water. Furthermore, we have investigated the Prandtl-number dependency of the turbulence quantities in thermal fields such as Nusselt number, temperature profile, temperature variance and eddy diffusivity for heat. Since the two-equation heat-transfer model provides information on temperature fluctuations, which cannot be obtained by the conventional model based on the assumption of the turbulent Prandtl number, the present model provides a powerful tool for critical design of thermal equipments.

View full abstract

We developed a reaction rate model of turbulent premixed flames to predict time-averaged profiles of the velocity, temperature and species concentrations in practical combustion systems based on the experimental observations of turbulent properties in a combustor. The fundamental features of the model consist of chemistry- and mixing-controlled reaction rates which vary with the local Damkohler number. The comparison of numerical predictions with experiments demonstrated the applicability of the concept to the flames of a wide range of Damkohler numbers.

View full abstract

An experimental study was made of the temperature fluctuation of a hydrogen-methane jet diffusion flame discharged normal to a free stream of air with a uniform velocity profile. The time-resolved measurement of temperature in the flame was carried out with the application of the Rayleigh scattering method. The mean temperature, root-mean-square (r. m. s.) temperature and probability density function (p. d. f.) of fluctuating temperature were obtained in the various transverse cross sections normal to the jet axis of the flame. The mean temperature profiles obtained by Rayleigh scattering method were compared with those measured by the thermocouple. The results showed that the high-temperature region is located below the jet axis. It was also found that a maximum r. m. s. temperature was located at the top edge of the flame, where the p. d. f. had two peaks.

View full abstract

An attempt was made to simulate the laminarization phenomena of strongly heated gas flows within a circular tube by means of a Reynolds stress turbulence model by Launder and Shima. Numerical results show that the adopted model can reproduce the streamwise variation of the Stanton number of laminarizing flows, while there remains the task of improving the prediction accuracy in the subtle stage of the turbulent-to-laminar critical region. It was also found that when the flow is laminarized due to strong heating, radial and peripheral components of normal Reynolds stress are preferentially attenuated and consequently / the inherent anisotropy of circular tube flows is amplified.

View full abstract

To develop a new type of mid-capacity transformer which uses perfluorocarbon liquid as the cooling and insulation medium, the basic flow and cooling characteristics in the windings were investigated. Three two-dimensional winding models having different horizontal and vertical duct spacings and number of coils in a section were used for the test. The models were composed of model conductors which had thermocouples and thin heater pins soldered and wrapped in insulating papers. The actual horizontal duct spacings, the horizontal and vertical temperature distributions of the windings and pressure drop were measured for each of the three models. The flow condition and winding configuration that cause overheating of the conductors were studied, and the influence of ripple or roughness of the insulating paper surface on pressure drop and heat transfer characteristics of the windings was investigated.

View full abstract

A hydrogen-burning small test combustor without air holes in a liner wall was designed on the basis of a concept as for lean combustion of hydrogen in a primary combustion zone under conditions of very strong swirl. After some preliminary work on the test combustor, a hydrogen-burning complete combustor for use in a hydrogen-fueled micro gas turbine was developed. The combustion test of the complete combustor showed that it was capable of maintaining very stable and efficient combustion over the required range of engine speeds and operation conditions. The combustor had the high rate of volumetric heat release of 3.5×10³MW/(m³·MPa), and attained high combustion efficiencies of over 99.95%. Hydrogen combustion in the combustor could result in low NO_x emission levels in comparison with hydrocarbon combustors.

View full abstract

The swirl-type injector used formerly as a single-point injector has faults, in that, it becomes difficult to generate a thin fuel film at a low fuel flow rate. A new type of single-point injector was proposed in this paper to solve the above problem and to improve its atomization performance. It is a impingement-type injector. Liquid fuel jets impinge on the solid wall, and the thin liquid film is issued from the injector. This type of injector has the following advantages. (1) The impingement-type injector can generate a thin liquid film even at a lower liquid flow rate. (2) The liquid film issued from the impingement-type injector is thinner and spread out more widely. The drops generated by the disintegration of the liquid film are finer. (3) The circumferential mass flux distribution of drops is more uniform. (4) The upwash generated by the impingement of the films promotes film spreading. However, a large upwash limits the uniform circumferential distribution of drop mass flux.

View full abstract

A fundamental test of running a very small wooden ship model with laser propulsion was performed. The pulsed Nd·YAG laser beam (λ=532nm) through a lens placed on a ship was focused to cause a laser-gas breakdown above it. The blast wave of hot air involving a shock wave and gas expansion by a plasma was used for propellant. Experimental results of the thrust and the efficiency of laser energy converted to propulsion force were F=4.0 [dyn] and η=7.4×10^-5[%], respectively.

View full abstract

A numerical simulation is performed on the deflection characteristics of the incident measuring light rays passing through the thermal boundary layer developing along an isothermally heated flat plate. The additional fringe order shift and fringe position displacement due to light deflection are calculated, and compensation for the deflection measurement uncertainty is mede by analyzing the fringes of real holographic interferograms. The validity of this compensation is confirmed by experiment on the combined free and forced convection flow field in the thermal entrance region of a horizontal rectangular duct with a constant wall temperature.

View full abstract

The k-ε model has been widely used to calculate turbulent flows in engine cylinder ; however, it has not been evaluated adequately whether the k-ε model can be applied to anisotropic in-cylinder turbulent flows. In this paper, several modified k-ε models and the algebraic stress equation model are evaluated for an anisotropic in-cylinder flow by a comparison of predicted velocity field with experimental result. Based on the comparison, a modified k-ε model is newly proposed which can describe the mean and turbulent velocities of the flow in the pancake-shaped combustion chamber. The modified k-ε model, however, still lacks the ability to reproduce turbulence in the combustion bowl in a piston.

View full abstract

The high-performance electric preciptation system used in power stations captures charged particulate by means of corona effect. Owing to the danger of high voltage and heavy equipment, the system is only able to be applied in stationary heat engines. Using a low-voltage filter system, experiments were performed in this study. Neutral particulate in the exhaust from a diesel engine passing through negative wire mesh will be negative-electron charged and captured by conductible positive carbon fiber. The findings show that the particulate concentration is reduced by 50% or so. Because there is the same or better safety performance that in the corona system. this low-voltage filter system could be applied to commercial diesel engines.

View full abstract