In this paper, numerical analysis of the entrance flow and the die swell of visco-elastic fluid in the dies are carried out. This method employs the GSMAC-method for the analysis of the momentum equation and the MUSCL-TVD method for the analysis of the constitutive equation. And we use larson model equation and the constitutive equation of visco-elastic fluids. In order to investigate the validity of this present analysis, the present solutions are compared with the experiment data of the HDPE-melts in the former literature. This present solutions of principal stress difference and swell ratio are in good agreement with the experiment data. And we investigate effects of flow rate and the slit length on die swell.

抄録全体を表示

The paper presents a new approach for unsteady coupled system of fluid and body motion of the compressiblie flows on a moving grid system. To assure geometric conservation laws, a finite-volume formulation in the space-time computational domain, of four demension for three-spacial dimensional geometry for example, is adoped in the scheme. The resultant fully implicit scheme is solved iteratively at every time step in order to assure the conservation laws and accuracy. The inner iteration is performed through the efficient and stable Rational Runge-Kutta scheme, so that the algorithm of the scheme can be treated explicitly notwithstanding the implicit scheme. This approach is just the same as pseudo-time concept. This paper gives an interpretation between inner iteration strategy and pseudo-time approach. The basic formulation is given mainly for the one-demensional space in detail in this paper, but the extention to multidimension is straightfoward. The basic feature of the present scheme is investigated and some examples of coupled interaction problemes of fluid and body motion, are given.

抄録全体を表示

Moving particle semi-implicit (MPS) method was developed for analyzing incomplessible flow with particle models. In the present study, a numerical model for thin walls is developed and the MPS method is extended to analyze fluid-structure interactions with large deformation of free surfaces and walls. First, sloshing in a rigid rectangular tank is calculated. A finite amplitude wave occurs in a shallow tank, while a standing wave appears in a deep tank. Results are in good agreement with the experiments by Hayama et al., though the calculated sloshing amplitude is smaller than the experimental one in the deep tank. This is attributed to unknown dissipation mechnisms in the present method. Next, sloshing in a tank with elastic walls is calculated. The calculated resonance peak is shifted to a lower frequency due to large deformation of the walls.

抄録全体を表示

This paper presents an application algorithm of CIVA method, which is so flexible as to enable to integrate the CIP, gridless and particle method, to incompressible fluid simulation. The CIVA-based algorithm that is combined with the methods such as the control volume method, the MAC method, the automatic unstructured mesh generation method makes it possible to simulate many kinds of problems having complex geometries and phenomena with a high accuracy. To confirm the possibility, numerical analyses for some benchmark problems were executed. The results showed that the method achieves high accuracy and has high flexibility even for the flows of high Reynolds number, complicated geometries, moving boundaies and free surface.

抄録全体を表示

A full three-dimensional numerical simulation of a flow in a rotating axial rotor was conducted for a flow rate of slightly smaller than stall onset. It is intended in this simulation to confirm that the novel mechanism of stall accompanying intermittent tube-like vortices found in the authors' previous simulation of a flow in a single inter-blade passage of a special rotor occurs also in a practical rotor. It turned out that the early stage of stall of a rotor is characterized by tube-like vortices and tube-like high pressure blobs of similar configuration flowing downstream with tube-like vortices. This finding suggests that widely accepted concept of stall onset that blockage of an inter-balde flow passage by two-dimensional separation due to high angle of attack is the key mechanism should be re-examined. The stall propagates to the inter-blade passage behind. Strength of stall at the rotor tip is not uniform circumferentially and this non-uniformity is found to rotate in the same direction as fully developed stall cell observed experimentally with almost same speed as the experiment relative to a rotor. The process in this simulation is expected to explain the formation of stall cell.

抄録全体を表示

A uniform transverse magnetic field is imposed on liquid metal MHD flows. Spatial development of Karman vortex between two insulator walls is numerically simulated. When a strong magnetic field is imposed, MHD diffusion reduces the vortices including three-dimensional component to quasi-two-dimensional ones in downstream. The condition for development of two-dimensional turbulence is discussed.

抄録全体を表示

There are several hypotheses as to why stenosis appears in an artery. However, there is no accepted theory as yet. In this study, axisy mmetric stenosis growth in an artery is modeled based on three hypotheses; (1) high wall shear stress theory (2) low wall shear stress theory (3) oscillatory shear index theory. In each hypothesis, stenosis growth is simulated numerically. Flow is assumed to be periodic, incompressible and axisymmetric. Non-Newtonian viscosity of blood and movement of arterial wall are considered. Flow pattern, concentration pattern and distribution of mass flux on the wall are computed. By using these results, stenosis growth process is obtained. It is found that the stenosis growth process based on the high wall shear stress theory is consistent with the clinical case. However, stenosis does not grow so much in the case of the low wall shear stress theory and the oscillatory shear index theory.

抄録全体を表示

An axisymmetric impinging jet onto a flat wall has studied using a numerical model of incompressible viscous steady flow, especially close to the impingement wall. The Navier-Stokes equations are solved in terms of stream function and vorticity with the combined upwind and central finite difference method of an alternating direction implicit way. Distributions of velocity and pressure of the impinging jet are calculated at several large Reynolds numbers using non-uniform grids in the axial direction of the jet for three kinds of fully developed velocity profile at the inlet of the impingement region. Calculated distributions of velocity and pressure agree well with experimental data of impinging air jets. More over, the effect of the inlet velocity on the flow and the flow structures of impinging jets are discussed.

抄録全体を表示

Light emission properties of xenon-water cavitation in circular orifice flows were investigated experimentally and theoretically. The emission was enhanced by dissolving a small amount of Xe gas into water, and its spacial pattern was taken by a chilled CCD camera and an image intensifier. A spectrum of cavitation emission was measured by a spectrometer with a photon counter. Based upon the collisional and radiative theory for plasmas and a two-temperature model, synthetic spectra were calculated for emission of a collapsing xenon bubble at the minimum radius. It was found that the light emission pulsated and was intense around the flow axis a little downstream of the orifice, and the experimental spectrum was predominated by continuum and its intensity distribution was very similar to the theoretical ones with electron temperature of about 12000 K to 14000 K. Mean emission energy per unit pulse was estimated experimentally to be smaller than 5×10^-9 μJ/nm·sr at the wavelength of 450 nm, which is consistent with the plasma conditions from the emission spectrum.

抄録全体を表示

Dynamics of a spherical bubble cloud with nuclei size distribution is investigated numerically when the surrounding pressure is decreased from 50 kPa to 10 kPa and then increased to 50 kPa stepwisely. A set of governing equations for the motion of bubble cloud are formulated combining the averaged equation and the equation of bubble motion with emphasis on the internal phenomena of each bubble. The nuclei size distribution has significant effects on the cloud dynamics in comparison with the results for mono-dispersed cloud. When the cloud contains multi-dispersed bubbles in their size, the maximum pressure in the bubbly mixture becomes lower than the case of mono-dispersed one. However, the maximum pressure emitted from each bubble has the same order with the mono-dispersed case. The range where high emitted pressure is observed is wider than the mono-dispersedcase.

抄録全体を表示

The flow over a forward facing step is very complicate having two vortex regions just before and on the step. In this paper, the flow characteristics over a forward facing step are made clear experimentally by the mean and fluctuating velocity measurements using a Laser Doppler Velocimeter (LDV) and numerically by κ-ε turbulence model. The control of causing of the vortex region on the upper step surface by putting a small obstacle on the lower step surface is also discussed. It is made clear that causing of the vortex region on the upper step surface can be supressed by the small obstacle and then the force exerted on the vertical suface of the step decreases.

抄録全体を表示

Reattachment length of the flow in tube downstream to contraction was investigated by experiments and numerical analysis. From the results of measurement by flow visualization technique, it was confirmed that the law of similarity for reattachment length with respect to the dimension of set up was held, while the upstream arrangement to contraction could vary attachment length. And the results also showed that there appeared three modes of flow reattachment according to the increase of Reynolds number, and that there existed the dependencies of reattachment length to contraction length and diameter ratio of contraction to the tube. Numerical analysis was made to evaluate reattachment length quantitatively for the mode corresponding to low Reynolds number, and the model expression for determining reattachment length was established.

抄録全体を表示

Unsteady non-uniform liquid film flow on a rotating disk. which starts to rotate impulsively from rest, is analyzed by asyptotic methods. A short-time-scale expansion for the transient film profile near the rotating axis is derived as a function of space and time. The effects of gravitational force and surface tension as well as centrifugal force are identified for a case in which the initial film thickness is even function of radial component and the non-uniformity in film thickness is much smaller than the film thickness itself. In this time scale, the centrifugal force has little effect on the spreading process at the very beginning of the stage due to the effect of the local inertia, while it becomes dominant as the angular velocity of the film develops. The gravitational force has smaller influence on film than the centrifugal force, but the rate of its effect takes the same order of magnitude through out this time scale. Although the surface tension has even smaller influence on the process, it prevents the liquid from spreading.

抄録全体を表示

A calibrator of laser Doppler velocimeters (LDVs) has been developed as the measurement standard of flow velocity. This calibrator has a rotor with single fine wire and generates various velocities from the rotation. To minimize the influence of wire deformation, the orbit diameter in movement was measured with an interferometer as the distance of a rotor traversing. A measurement control volume of tested LDV was used to detect the wire orbit position. The calibration factor of tested LDV was decided as a ratio of wire velocity to the average frequency of the Doppler burst signal. In this paper, the uncertainty of LDV calibration with this calibrator was analyzed and unknown sources were quantified by the experiment. The maximum uncertainty source was pitching of the carriage table, which would be reduced by remodeling of the devices. The combined standard uncertainty of the calibration was about 100 ppm in the velocity range of 2 to 30 m/.

抄録全体を表示

Nowadays, a non-intrusive measurement tool of thermodynamic variables with high sensitivity is strongly demanded for analyses of highly rarefied gas flows. REMPI (resonantly enhanced multi-photon ionization) is the most suitable technique for measurement of gas molecules with very low density. In this study, to examine the fundamental properties of the REMPI spectra, 2R+2N₂-REMPI spectra including the spectral broadening are calculated and compared with the experimental results. From the calculated REMPI spectra, spectral lines adequate to measure temperature are proposed, especially at relatively high temperature where the measurement error becomes larger because of the overlap of the spectral lines.

抄録全体を表示

Experimental study to generate T-S waves and oblique waves in a flat plate boundary layer is carried out. It is regarded as the first step towards controlling the boundary-layer transition. In order to generate T-S waves and oblique waves, an array of piezoelectrically driven actuator pieces attached to the flat plate surface is used. The actuator is thin enough to keep the boundary layer in the laminar-state. In order to generate various oblique waves, the actuator consists of an array of 10 small pieces aligned in the spanwise direction. Each piece can be controlled separately. It is shown that various oblique waves can be successfully generated using this actuator. It is also shown that when the three-dimensionalization mode is induced, the onset of turbulence takes place locally at a certain spanwise location.

抄録全体を表示

Pumping characteristics and power consumption of the screw vacuum pump with two piecewise constant lead angles were investigated. Characteristics of the screw vacuum pump with uniform lead angle (isochoric design) had already been analysed by the balance among geometrical pumping speed, net throughput and leaks. We took account of varying of transfer volume and clearances, and modified the analytical model to be applicable to the two piecewise constant lead angle screws. Experiments were carried out by a screw vacuum pump with two piecewise constant lead angles. The ratio of inlet side lead angle to outlet side one of the pump was 0.53. The power consumption of the pump was reduced by 35 percent in comparison with the pump of isochoric design. The analytical model simulated both pumping speed and power consumption with good accuracy.

抄録全体を表示

To investigate natural convection in a cylindrical cavity with sinusoidal gravity modulation, the authors calculated it by the finite difference method. The full Navier-Stokes equations in axisymmetric cylindrical coordinates are simplified with the Boussinesq approximation. The Rayleigh number R_a and the Prandtl number Pr are equal to 2.8×10⁵ and 0.7, respectively. While except for a few cases, non-dimensional gravity acceleration amplitude η is fixed at 0.1, non-dimensional angular frequency ω is varied from 10⁰ to 10⁵. In the tested range, velocities and temperatures at any points fluctuate sinusoidaly at the same frequency as the gravity acceleration does, while their amplitudes are varied gainst ω. Temperature's amplitudes are much smaller than velocioy's ones, everywhere. It is made clear that the velocity's amplitude can have a maximum value at a certain angular frequency ω_max although any instantaneous flow patterns are similar to that without gravity modulation. More specifically, globally ω_max≒200, but ω_max varies form 10¹ to 10³ depending on the location. In addition, further results with much larger acceleration amplitudes are presented at ω=200.

抄録全体を表示

A water-fine particle suspension with a uniform initial temperature and concentration in a rectangular cell was heated from a vertical wall and cooled from the opposing vertical wall. The dispersed particles had a distribution in size. In this natural convection, many layers separated by almost-horizontal sharp interfaces were observed. In the beginning the many layers appear, and each interface of the layers falls gradually with a constant velocity, and finally all layers vanish. To make this phenomenon clear, concentrations and mean diameters of the particles in each layer and temperature distributions were measured. Moreover natural convection of the other suspension with the particles which had a narrow size distribution was also investigated.

抄録全体を表示

This paper presents a non-contact AC optical technique for measuring the thermal diffusivity of a wafer-like thin solid sample. The technique is based on heating one solid surface with a odulated laser beam and monitoring the corresponding temperature modulation of the other surface across the sample using the reflectance of a probe beam. The phase lag between the two laser signals is independent of the optical properties of the sample as well as the laser intensities, and can be related to the thermal diffusivity. A detailed theoretical analysis is presented to estimate the thermal diffusivity of the solid from the phase lag measured as a function of the modulation frequency. A closed form solution in the high-frequency limit is derived in order to provide a simple estimation procedure. The effect of misalignment of the two lasers is studied and conditions for making robust measurements are suggested. As a benchmark of this technique, the thermal conductivity of a single crystal silicon sample was measured to within 2 percent of reported values.

抄録全体を表示

The present paper describes the vibration effect of impingement surface on two-dimensional impingement heat transfer with confined wall. The local temperature distribution on an impingement surface was measured using a thermosensitive liquid crystal sheet and an image processor. The net heat flux was obtained by solving the three-dimensional heat conduction equation in the heated section. Experiment was carried out by using air as a working fluid. Experimental conditions were as follows; Reynolds number Re=1000∼10000, dimensionless distance between nozzle and impingement surface h/B=1.0∼4.0, frequency f=0∼100 Hz and amplitude a=0.5, 1.0 mm. Local Nusselt number was improved for comparatively low Reynolds number and low frequency and was depressed for high frequency. In the case of heat transfer enhancement, vortices on the impingement surface were renewed frequently, and on the other hand, in the case of heat transfer depression, thermal boundary layer thickness increased in appearance by the vibration of impingement surface.

抄録全体を表示

The effects of spacing between dual jets on the heat transfer characteristics and flow behaviors over the plate surface were investigated experimentally when the dual slot jets impinge on the target plate. The local heat transfer distributions were measured for various spacings at small nozzle-to-plate separation distances and the flow pattern also visualized by the oil-titanium IV oxide method. An infrared radiometer with a two-dimensional array of InSb sensor was employed to clarify space varying in the flow and heat transfer accompanying by the interferences between jets, and jet and spent flow. The characteristics of heat transfer, flow patterns and thermal distributions change with the spacing of the dual jets and the nozzle-to-plate separation. The phenomenon of axes switching, which is caused by the differences in the self-induced velocity in the non-circular vortices, was observed farther downstream compared with that of a single impingement jet. These phenomena consequently played an important role in the heat transfer enhancement.

抄録全体を表示

The two-dimensional, steady, pool film boiling heat transfer from a horizontal circular plate facing downward to a stagnant saturated liquid is studied theoretically on the basis of the assumption that the liquid-vapor interface is smooth and radiation can be disregarded. The relevant governing equations for the vapor film are solved for water at atmospheric pressure using two-equation boundary-layer integral method. It is shown that the dimensionless temperature profile is affected by the wall superheat ΔT_sat and that a ratio of Nu to X^0.2 an increasing function of ΔT_sat. Here, Nu represents the mean Nusselt number and X the film-boiling Rayleigh number. In addition, it is revealed that one-equation boundary-layer integral method developed by Nishio et al. is fairly accurate in predicting the film thickness, the representative radial velocity and the mean Nusselt number.

抄録全体を表示

Forced convection boiling of subcooled water was performed in a horizontal rectangular channel with heated surface on the bottom. The experiment was conducted for heating surfaces, 10 mm, 20 mm and 40 mm in length. Microbubble Emission Boiling, MEB, was observed in subcooled transition boiling and easy to be generated for the shorter heating surfaces. In higher flow velocity of subcooled water, MEB was generated at even lower subcooling. Stormy MEB was observed at both the higher subcooling and the higher flow velocity of water. In the stormy MEB, the heat flux rose up rapidly above CHF (Critical Heat Flux) with larger acoustic noise and vibration.

抄録全体を表示

An experimental study was performed to investigate the local heat transfer around a cube mounted on the wall. The Reynolds number ranged from 4.2×10³ to 3.3×10⁴. The cube lied in the turbulent boundary layer. The surface temperature distributions around the cube were measured with thermocouples under the condition of a constant heat flux. The local heat transfer is very high in the region of horseshoe vortex formed around the cube. On the top and side faces of the cube, the heat transfer has a maximum at the reattachment region of the shear layers separated from the edges of front face of the cube. The average Nusselt number on each face of the cube is given as a function of Reynolds number. The overall Nusselt number of the cube is expressed by Nu_m=0.38Re^0.60.

抄録全体を表示

A numerical analysis based on the finite difference method is made on a three-dimensional unsteady separated and reattached flow around a surface-mounted hexahedron in a channel with two parallel walls, and the heat transfer characteristics therein are also analyzed. Reynolds number treated is 1000 and Prandtl number 0.7, respectively. At Re=1000, the horseshoe vortex becomes unsteady in the vicinity of the side surface, and the shear layer separated from the front corner rolls up and the vortex is shed to the downstream. Futher, it is clarified that the vortices are periodically shed from the side corner. It is also shown that on the side, top and rear surfaces, the local Nusselt number varies with time originating from the vortex shedding, but the local Nusselt number on the front surface exhibits no essential variation with time.

抄録全体を表示

Several direct mumerical simulations (DNS's) have been performed for the turbulent heat transfer in a channel flow. The coupling of the velocity and thermal fluctuations causes a turbulent heat flux. In the present study, the effect of a difference in a thermal boundary condition is examined. Two DNS's were carried out for constant wall heat flux and constant wall temperature difference conditions with Pr=0.71 and 0.025 to obtain statistical quantities such as mean temperature, temperature variance, turbulent heat flux and their budget terms. A mixing length approximation is applied to analyse the relation between the mean temperature and the turbulent heat flux profiles. The effect of difference in the thermal boundary condition is discussed.

抄録全体を表示

A rarefied gas flow in a porous medium has been numerically investigated for the slip-flow regime, using a two-dimensional periodic model of porous structure. A macrmoscopically uniform flow is assumed to pass through a collection of square rods placed regularly in an infinite space, where a temperature gradient is imposed perpendicularly to the flow direction. Continuity, momentum and energy equations are solved numerically, using the slip velocity and temperature boundary conditions, to describe the microscopic velocity and temperature fields at a pore scale. The numerical results thus obtained are integrated over a unit structure to determine the effective permeability and the molecular diffusion due to tortuosity. The effects of Knudsen number on the permeability and thermal conductivity are elucidated

抄録全体を表示

A new thermosyphon for the use in the non-gravity field has been proposed by using a magnetic fluid as a working fluid. The heat transport of the thermosyphon is driven by a magnetic force instead of the gravity force. In order to avoid too much concentration of the magnetic fluid in a heating region, a liquid bypass line is installed between a heating region and a cooling region so that the highly concentrated magnetic fluid flows from the heating region to the cooling region. The perforance of the thermosyphon has been evaluated based on an analytical estimate of total pressure rise and loss through the magnetic fluid loop. Under the conditions of system pressure 0.02 MPa and the fluid temperature 60°C, it has been confirmed that the thermosyphon with the working fluid of a water-base magnetic fluid has the heat transport capacity of 100 W with the circulation force of 3500 Pa.

抄録全体を表示

In order to discriminate definitely between huge waves and disturbance waves appearing in the flows with liquid visosity of ν_l=1, 10 and 20cSt, a cluster analysis by K-mean algorithm was applied using wave width, wave velocity and wave height determined by wave-vein analysis. Appearance flow conditions of these waves were presented and flow patterns such as plug flow, huge wave flow and annular flow were determined. It is clarified that the appearance region of waves gets narrow with increasing liquid viscosity. Comparison of the present results with the flow map proposed by Furukawa et al. was made and it is clarified that froth-annular flow is the region of coexisting huge waves and disturbance waves. Wave width, wave velocity and wave height of huge wave and disturbance wave were determined and the effects of liquid viscosity on these flow parameters were found out.

抄録全体を表示

Effect of narrow band nonuniformity on unsteady heat up of water vapor under radiation-conduction combined heat trannsfer is examined by comparing the results of numerical simulations with and without incorporation of narrow band nonuniformity. Results of examination exhibited that the contribution of radiative heat transfer to the heat up rate of water vapor may be overestimated up to almost twice, if the narrow band nonuniformity effects is neglected. Separate analyses of radiative energy attributed to wall emission and gas emission clarified that the absorption of wall emission is overestimated and, on the contrary, he absorption of radiation energy emitted by water vapor itself is underestimated if the narrow band nonuniformity is neglected. It is also clarified that the negligence of the narrow band nonuniformity give rise to severe error in the estimation of absorption rate of wall and gas emission even in the case where the negligence of narrow band nonuniformity bring little change to the temperature distribution.

抄録全体を表示

The main interest was taken in the effects of the density distribution on the radiation transfer in the scattering media. The frost layer in the early growth stage under the control of diffusion was chosen as the media. Numerical analysis of the radiation transfer was carried out by using a modified Monte Carlo method on the basis of the geometrical optics. A diffusion limited aggregation theory was introduced to simulate the growth of frost layer with treelike structure. The relation between the density distribution and the transmittance of the simulated model was made clear by the numerical analysis. On the basis of its result, the optical measurements in the visible wavelength region were utilized to evaluate the density distribution in the frost layer, which varied with the cooling condition, from the transmittance values.

抄録全体を表示

This paper proposes a dynamic equation of state for the partially frozen zone between the freezing front and the ice-lens forming front, i.e. the so-called frozen fringe, in a one-dimensionally freezing porous medium saturated with liquid. The dynamic equation of state is derived from an enthalpy balance of the liquid flow in the frozen fringe, and it can be used to quantify the segregation ratio, which represents the segregating volume in the unit volume of a porous medium for a unit time. In addition, the study has clarified that the frozen fringe consists of two zones, namely a sub-cooled zone including the ice-lens forming front where segregation occurs, and an equilibrium zone including the freezing front where in-situ freeze occurs. In the sub-cooled zone, the segregation produces an expansion strain, which was observed by a frost heave experiment. Thus the proposed theory is qualitatively verified by the experimental results.

抄録全体を表示

Physical insight is cast to the experimental documentation that a flame propagates at high speed within a vortex tube filled with a flammable gas mixture, A prototype model used for this purpose is the Rankine type vortex with the core region and free-vortex regions being occupied by the flammable gas and an inert gas, respectively. The infinite vortex is ignited at the central cross section so that two flame surfaces propagate in the opposite directions. It is shown that the spiraling of axial vortex filaments caused by the presence of flame (volumetric source) generates the azimuthal vorticity distributions at the shoulders of the flames, which induces the axial velocity at the flame tip thus enhancing the flame propagation speed proportional to the vortex strength. An explicit expression for steady flame tip propagation speed is derived in a general frame work from the requirement of single-valued pressure field on account of three-dimensional flow structure, which yields good agreement with experimental data. The analysis is also applied for other related problems of interest.

抄録全体を表示

Commpressible Euler equations are numerically solved coupled with the fuel and energy conservation equations which include source terms due to an overal first-order one-step reaction, in order to examine the temporal evolution of flow after a flammable gas mixture confined in an infinitely long Rankine vortex core is ignited at a central part. The distributions of axial vorticity, azimutjhal vorticity and axial velocity clearly show the convective draft of flame tip by the axial velocity induced by the azimuthal vorticity distribution generated on the shoulder part of the curved flame surface due to the spiraling of axial vortex filaments by the thermal expansion effect of flame, thus vailfying the flame propagation mechanism and predictive expressions proposed in the first report.

抄録全体を表示

By using a cyclone-jet combustor, we have investigated premixed flames in a highly turbulent flow over a wide range of turbulent properties, with u'/S_L exceeding 10 in a stationary jet flame. With Mie scattering, we obtained tomographic images to visualize the flame structure. We have investigated the velocity field in cold flow with a hot wire anemometer. Results show that, as the turbulence is relatively low, the scale of the turbulence is large and one flame sheet wrinkles, which shows that a typical wrinkled laminar flame is formed. As the exit velocity is further increased, the flame wrinkling is increased, but the size of its wrinkles becomes small with many cusps of the flame front. In some cases, the burned gas may exist in the unburned gas, and vice versa. When the velocity exceeds 30 m/s, one continuous flame sheet cannot be sustained, and the destruction of laminar flame structure occurs, in which small reacting eddies in an intermediate stage of combustion may be embedded. These results are supported by the fact in previous experiments that the probability of intermediate temperature between those of unburned and burned gases becomes high and the PDF of temperature is no longer bimodal. Thus, it appears that, a flame structure changes to be that with distributed reaction zone as the intensity of turbulence is increased.

抄録全体を表示

The Lewis numbers of lean heavy-hydrocarbon fuels are larger than unity, and hence, their flames are prone to extinction in a shear flow, which occurs in a turbulent combustion. Here, propane is used as a representative fuel of heavy-hydrocarbon fuels because the Lewis number of lean propane/air mixtures is larger than unity, and an attempt to improve its combustion characteristics by hydrogen addition has been made. A tubular flame burner is used to evaluate its improvement, since a rotating, streched vortex flow is established in the burner. The results show that with hydrogen additon, the fuel concentration, the flame diameter and the flame temperatuire at extinction are reduced and its combustion characteristics are imporved. However, it is found that the effective equivalence ration at extinction cannot become so small as that of lean methane/air mixture, which has a Lewis number less than unity.

抄録全体を表示

Experimental and analytical study was performed for surface reaction model that was used for numerical simulation of pulverized coal firing boiler. Combustion efficiency in downstream of premixed flames was examined for a subbituminous coal, two bituminous coals and an anthracite. The influence of SR (stoichiometric ratio), combustion temperature, FR (fuel ratio) and initial particle diameter was examined. Reaction rates of oxidation and gasification, and surface area model of coal char were examined by comparison between experimental and analytical results. The outer surface area was assumed not to be changed during devolatilization, but to be reduced by surface reaction. Effective surface area for reaction was assumed to be decreased when ash content in char was increased. By using this surface area model, comparison of the predicted combustion efficiency with the experimental data showed good agreement for SR 0.5∼1.5. Reaction rates were decreased when FR of coal increased. The effect of coal type was significant for gasification reaction. The activation energy of gasification was smaller than the previous study that was obtained without oxygen. The reaction yield by gasification at SR 0.8 was about 10%, gasification was found to be important for numerical simulation of pulverized coal firing boiler.

抄録全体を表示

This study tried to reduce NO_x and particulate emissions simutaneously in a high speed direct-injection diesel engine. Fuel-rich combustion was used to reduce NO_x emission at initial combustion stage and high turbulence combustion was used to reduce particulate emission at diffusion combustion stage. Experiments were carried out to investigate the effects of high squish combustion chambers with small throat on combustion process and exhaust emissions. Results showed that in the high squish combustion chamber, both NO_x and particulate emissions reduced with retarded injection timing. Endoscopic high speed photography was employed to analyze experimental results. In the original combustion chamber at the early combustion stage, the high luminosity flame tended to spread out of the cavity. Fuel-rich mixture burned slowly over the piston crown. In the high squish combustion chamber, flame rotated violently around the bowl. Fuel-rich mixture was kept beneath the squish lip, and was mainly oxidized inside the bowl. This combustion phenomenon was considered to be one of the reasons for low particulate emission in high squish combustion chambers.

抄録全体を表示

In-Cylinder gas flow is one of the most important factors that affect to the combustibility of premixture in the cylinder and heat transfer to the cylinder walls of an engine. Many papers have been published about gas flow measrement in an engine and flow visualization method is one of the effective means. In our previous report, a method of three dimentional measurement, that is called "color layer method", was introduced that the depth information perpendicular to the image plane is represented by color of tracer. In this paper, improvements of measurering system, mainly about the optical systems, were made and examined measuring accuracy. Flow velocity in a tube of laminar flow and that of unsteady flow with vortex were measured and they were compared with theoretical value and measured value by LDV respectively. Furthermore, flow in a model engine was measured. As a result, it is confirmed that the system was improved in accuracy and was able to adopted to unsteady flow measurement.

抄録全体を表示

Early stage fuel injection and its combustion system with lean and homogeneous mixture under compression ignition have been investigated to decrease the pollutant emission of DI diesel engines. This method enables drastic reductions in both NO_x and smoke emissions, but the operating load range is restricted due to the unusual combustion phenomena such as unsteady combustion or knocking. This research investigated the possibility of broadening the operating load range and reducing unburnt HC using Dimethyl Ether (DME) premixed gas charged through the intake pipe. However, the broadening in load range was found to be less than BMEP 0.1 MPa even when compression ratios were reduced and high octane gaseous fuel was mixed. Therefore, a DME premixed charge could be used only at light loads. At heavier loads, a hybrid fuel system with the DME premixed charge and diesel fuel injection is necessary. It was found that DME operation yielded significant reductions in both NO_x and smoke, and hybrid fuel operation could reduce smoke emission when DME was charged up to the knocking limit. Component analysis of unburnt emissions was made by gas-chrmomatography, and it was found that the high molecular weight fraction is reduced by the charging with DME mixture.

抄録全体を表示

A new diesel combustion concept termed MULDIC (MULtiple stage DIesel Cobustion), which can reduce NO_x emissions at high load conditions, was studied by means of engine test, combustion observation and numerical simulation. The engine test showed that simultaneous reduction of NO_x and smoke could be obtained with MULDIC operation. In-cylinder combustion observation was carried out in order to clarify the cause of the smoke reduction of the MULDIC. In these photographs, a first stage combustion was observed with nonluminous flame. In a second stage comustion, a highly sooting luminous flame was observed because the ignition delay was short. However this luminous flame disappeared quickly. Computational results showed that, oxygen is distributed evenly in the cylinder in later portion of the second stage combustion. Moreover, since cylinder temperature was very high, most of the soot oxidized rapidly. Thus, low smoke combustion was obtained.

抄録全体を表示

The combined cycle system becomes more complex in accordance with the gas turbine firing temperature. Recently, the steam cooled combined cycle was proposed as one of the advanced combined cycle, which was expected over 52 % (HHV) thermal efficiency. These types of combined cycle have many design parameters. And the optimum design method is necessary to determine such many parameters effectively. This paper presents an optimum design method for combined cycle, using non-linear programming. The calculation results show the method can be applied to the investigation of steam cooled combined cycle and it can be also useful for the effective design of the combined cycle. The results also propose an optimum system which is robust for various gas turbine design conditions.

抄録全体を表示