In this thesis, we propose the high speed and high acuraccy scheme for iscompressible fluid I analysis. The advection term in the Navier-Stokes equation has a nonlinear form and is therefore 1 difficult to analyze with accuracy. In this scheme, Navier-Stokes equation is divided into the advection term and the non-advection term, and the advection term is analyzed with a biquadratic element. The non-advection term is conventionally solved using the bilinear element. Using biquadratic element makes the calculation time for the coefficient matrices increase. In order to reduce the calculation time, the coefficient matrices are calculated with one point integration combined with the hourglass matrices, resulting in the high speed technique of GSMAC-FEM. This scheme is varified by six models such as the forced advection model in a cavity and the rotating cone model. These results showed that it could be analyzed precisely toward the two-and three-dimensional unstable flow.

View full abstract

The present paper treats a theoretical remark of the uniqueness of solutions in vortex methods. In the vortex method, the panel method is used to solve the Euler equations by taking account of the Kelvin theorem. However, it is an open problem whether the Kelvin theorem is necessary or not in the case where a bluff body impulsively starts in fluid at rest, because the initial flow does not have circulation. Furthermore, in three dimensional flows it is also an open problem how to use the Kelvin theorem for flows around bluff bodies. The present paper analyses two and three dimensional flows governed by the Stokes equations and shows that an additional condition such like the Kelvin theorem is necessary in two dimensional flows but in three dimensional flows it is not necessary to get a unique solution.

View full abstract

A method for controlling grid distribution for elliptic system of grid generation is proposed and developed. The elliptic grid generation method developed by Thompson utilizes source terms as control functions for controlling grid. In this paper, roles of the source terms of the elliptic grid generation equations are analysed in detail based on geometrical interpretation. It is shown that the control functions can be divided into three types of effects : direct grid-spacing ratio, non-orthogonality and mesh curvature. Based on the analysis, the paper gives a method that can control grid distribution as desired. Illustrative examples show a promising feature of the present method.

View full abstract

A solution-adaptively moving-grid method for unsteady shocked flows is presented in this paper. An elliptic system for solution-adaptive gridding and a finite-volume method formulated on space-time control volume for compressible flow solution are combined through an inner iteration process and solved simultaneously at every time step. The method is derived for two-dimensional space and validated using simple test problems. Application to a cylindrical implosion problem showed a promising feature of the method.

View full abstract

The intensity and the directivity of the aerodynamic sound generated from a plate having semicircular edges are investigated by wind tunnel experiments and numerical analysis. From the experiments, it is indicated that the intensity of the aerodynamic sound isn't proportional to U⁶ and the dipole-type directivity of the aerodynamic sound is biased toward the upper stream side with the increase of c/λ, where U, c, and λ are the wind velocity, the plate chord-length, and the acoustic wave-length, respectively. Numerical analyses for various thickness plates are performed in order to be compared with the experimental data. A combined method which consists of a two-dimensional computational fluid dynamics (CFD) code and a two-dimensional aeroacoustic code based on the Lighthill's equation is used. As a result, the calculated sound fields are qualitatively in reasonable agreement with the experimental ones. Moreover, the identification of the source locations are performed by the numerical method. It is found that the Lighthill's sound source term becomes large near the reattachment of a separation bubble and the trailing-edge. The latter source is the main contributor in the radiated sound field because the former is cancelled by the mechanism based on phase variations.

View full abstract

The thermal effect on a liquid film of uniform thickness on a rotating disk is analyzed by numerical method. The Navier-Stokes equations and the energy conservation equation in self-similar form for a case in which the disk temperature remains constant and equal neither the temperature of the surrounding gas nor that of the initial film, are solved by a finite-difference scheme. Assuming Newton's cooling law at the liquid-gas interface, the film thinning process was simulated for any Prandtl numbers. The numerical results reveal that if the Reynolds number is small and the product of Reynolds number and Prandtl number is up to unit order, the analytical expression in the lubrication limit gives a reasonable prediction for the transient film thickness. Furthermore, above calculations were extended by solving coupled equations for both liquid and gas phase with applicable boundary conditions, and the transient value of temperature and film thickness were obtained.

View full abstract

The effect of pulsatile flow conditions on vortex behavior downstream of stenosed was studied experimentally by means of PTV methods. Phase mean velocity vectors in the downstream region of stenosed model with 89% area reduction were measured under the condition of sinusoidal pulsatile flow of mean Reynolds number Re= 100∼300 and Stokes number St =151∼410. It was observed that, in all conditions, vortex ring was produced with a trailing jet downstream of a stenosed during the period of increase in flow rate. In the case of higher St number, the second vortex ring was produced in contiguity with the leading vortex ring. In the case of higher Re number, a row of small vortex rings was produced along the shear layer of the trailing jet. These vortex rings were propagated downstream field with constant velocity, and breakdown was followed. The position of vortex ring breakdown was only influenced by St number, and the ratio of propagation velocity of vortex ring to mean velocity in a period at the stenosed was constant under the all conditions.

View full abstract

Transitions of flow in periodically contracted-expanded channel and pressure characteristics are numerically investigated for effects of the channel geometries by using finite difference method. The flow fields are assumed to be two-dimensional and periodically fully developed. Steady state solution of the flow is obtained for relatively small Reynolds number, whereas it becomes unstable and self-sustained oscillatory one at the critical Reynolds number depending on the channel geometry. The critical Reynolds numbers are obtained for various channel geometries. It is found that the self-sustained oscillatory flow occurs as a result of Hopf bifurcation driven by a Tollmien-Schlichting wave. Moreover, it is seen that the pressure drop decreases for the expanded channel geometry compared with that of parallel plate flow for the subcritical Reynolds numbers, while increase for the supercritical Reynolds numbers. On the other hand, the pressure drop for the contracted channel is always larger than that of the parallel plate flow.

View full abstract

The vortex structure within a separated flow region changes greatly when the fluctuation is introduced in a mean-flow. This change was investigated by experiment. As the disturbance fluctuation, the periodic velocity fluctuation was introduced into the mean-flow using the disturbance generator at the upstream of separation shear layer. Frequencies of fluctuation were chosen based on the characteristic frequency of the separation flow. This characteristic frequency is measured in the shear layer or near the trailing edge without the disturbance fluctuation. It was found that the separation region was reduced most effectively when the fluctuation frequency is the same as the vortex shedding frequency of the separation flow without the disturbance fluctuation.

View full abstract

It is important to bio-fluidmechanics to understand the interaction among deformable surface, its surrounding flow field and hydrodynamical forces. In order to clarify this interaction, we have been using cylinder models having an elasto-flexible surface in uniform water flow. The laser light sheet method was used to visualize the deformation of the surface and the flow field around it. The experimental result showed that there were three different kinds of traveling wave generated on the surface, namely, a continuously-occurring wave, a periodically-occurring wave and a randomly occurring wave. In the case of the periodical and continuous waves, the frequencies of the waves were 6 to 9 times higher than that of the Karman vortex pair shedding from the rigid cylinder. Their wavelengths were about 1/8 compared with the theoretical wavelength of the Karman vortex street behind the rigid cylinder. It is indicated that generation of the wave was not caused by large pressure variation from the vortex shedding. The visualization results suggest that local pressure decrease flow separation from the surface and by the generation and moement of the small vortex were responsible for the apperarance and the propagation of the transient wave.

View full abstract

Three-dimensional flow structure induced by an elastic plate vibrating with a small amplitude has been studied experimentally. We focused on the effect of frequency of the vibrating plate praticularly. In this experiment, we first visualized the flow around the vibrating plate using tracer method with a laser light sheet. Next the velocity field near the free end has been measured using laser doppler velocimeter. The results showed that the jet-like flow was discharged from the center of free end and the discharged air was mainly sucked from side of plate. The motion of fluid around the plate seemed to be influenced by not only centrifugal force caused by the elastic vibration but also by pressure distribution related to vortexes generation. A change in frequency did not make marked differences in structure of flow, except for the velocity profiles of the jet induced near the free end.

View full abstract

We measured the characteristics (density, drag coefficient, magnetization) of a flowing ferromagnetic floc in a magnetic field. The purpose of this analysis is to make a calculation model for analyzing the behavior of the magnetic floc in a distorted magnetic field in a fluid flow. In the present experiments, we used kaolin suspended in water as influent and generated coagulated floc (containing suspended kaolin particles and seeded magnetic particles) . The trajectories of the floc in the magnetic field and fluid flow were experimentally observed by camera, and a calculation model of the trajectory was developed. The calculated trajectories from this model agree well with observed trajectories of the floc.

View full abstract

Cleaning method of semi-conductor device is developed applying cavitation jet which is generated through two different pressure nozzles. Its remarkable features are non-contact type, big pressure amplification ratio, wide cleaning effect area, having ultrasonic wave effect, easy introduction of functioned water in accordance with its purpose. Therefore, it is optimum cleaning for semi-conductor device which is mechanically weak and has an electrostatic charging problem. Comparison tests result in that cleaning efficiency of caviation cleaning is better than that of water jet cleaning, and cleaning time in case of cavitation jet is shorter than that in case of water jet.

View full abstract

The purpose of the present study was to develop a new snow fence with the airfoil snow plates for the prevention of snowdrifts and improvement of visibility on roadways. The design was in accordance with aerodynamic principles, and its performance was evaluated in a wind tunnel test in which natural snow was used to simulate blowing-snow. In general, the snowdrifts and visibility blockage behind the snow fence was caused by the alternate rolling-up of the bottom gap flow and the shear layer separated from the top of snow fence. It was found that the alternate rolling-up of the bottom gap flow and the shear layer separated from the top of snow fence could be prevented by using airfoil snow plates. On the basis of this finding, the airfoil snow plates with an optimum configuration were designed for the prevention of alternate rolling-up of their flows. Consequently, the developed new type snow fence with the airfoil snow plates plates had a high performance in preventing snowdrifts and improving visibility in comparison with the conventional snow fence.

View full abstract

The present study was to develop a new snow fence with the airfoil snow plates for the prevention of blowing-snow disasters. In the previous report, it was found that the developed new type snow fence with the airfoil snow plates had a high performance in preventing snowdrifts and improving visibility in comparison with the conventional snow fence. In this report, the characteristics of aerodynamic fluid forces and the flow-induced vibrations of the airfoil snow plates was investigated. Important results from the present study were that (i) the fluctuating fluid forces acting on the airfoil snow plates which set up the topmost part of the snow fence was the most values in all of them, (ii) the flow-induced vibrations in the airfoil snow plate attached on the topmost part was generated, and then become divergent, (iii) the fluctuating fluid forces and the flow-induced vibrations of the airfoil snow plate were completely suppressed when a perforated plate was attached at the top part of the snow fence.

View full abstract

We have investigated the rheological properties and the orientational distributions cof particles of a dilute colloidal dispersion, which is composed of ferromagnetic spherocylinder particles, subject to a simple shear flow. First, the governing equation of the orientational distribution function has been derived from the balance of the torques acting on the particle in an applied magnetic field. Then, the governing equation has been solved aproximately by Galerkin's method, after the solution was expanded in spherical harmonics. The results obtained here are summarized as follows. The orientational distribution function has a sharper peak for a stronger magnetic field, and the direction of the peask changes from a flow direction to a magnetic field one as the magnetic field comes to govern the shear flow. Since the direction of the particle is highly limited in the field direction as the magnetic field becomes strong, the viscosity increases significantly. The particle of a larger aspect ratio leads to the larger increment in the viscosity, since such elongated particles induce larger resistance in a flow field.

View full abstract

In this study, a three-dimensional unsteady numerical analysis has been made for a one unit of single row plate-finned tube heat exchanger located in a uniform flow. The effects of Reynolds number and fin pitch on local and mean Nusselt number and pressure coefficient were examined parametrically. The unsteadiness of the flow revealed some important effects on heat transfer from the fin-and-tube surface. The unsteadiness depends on not only tube diameter but also fin pitch. Karman vortex structure is observed in the instantanous flow field and twin vortex structure is seen in the time mean flow field. As Reynolds number or fin pitch increases, unsteadiness is strengthened so that low heat transfer area of the fin behind the tube decreases. Pressure coefficient shows a peak that would be related to increasing of the momentum by the unsteady flow.

View full abstract

Boiling heat transfer and critical heat flux characteristics were experimentally investigated in a SUS 316-vertical tube with a teflon-coated wire coil of varying wire diameter and pitch using HCFC -123 as the working fluid. The tube wall was uniformly heated by a direct electric current. The experimental conditions were mass flux G=200∼400 kg/m²s and the thermal equilibrium qualities at the exit of the tube x_eq =0.1∼0.8. The values of the boiling heat transfer coefficients in the tube with a wire coil were similar to those in the smooth tube in a nucleate boiling region, but in the region of forced convective evaporation a little enhancement of the heat transfer was observed. The critical heat fluxes in the tube with a wire coil at higher quality region were larger than those in the smooth tube.

View full abstract

The present paper describes the upward motion, the interaction and the impingement heat transfer to the upper wall of two thermal plumes from two heated sections on a horizontal plate. Numerical analysis was performed for dimensionless pitch P_i=0.26, 0.50, 0.76 and 1.0, Grashof number Gr=10³∼10⁶ and Pr=170, Numerical results show that the upward motion is divided into three patterns depending on the pitch of two heated sections. The first pattern is the unification of two thermal plumes like one for short pitch. The second one is the separate upward motion after the unification. The third one is the independent motion for long pitch. These patterns were investigated by visualization of thermosensitive liquid crystal suspension. The impingement local heat transfer on the upper wall is evaluated transiently for various Grashof numbers. Mean Nusselt number depends on Gr^-0.16.

View full abstract

The objective of the study is to clarify the effect of vapor velocity on Marangoni condensation characteristics for water-ethanol bainary vapor mixture. The specially desingned heat transfer block having trapezoidal sectional area to concentrate the heat flux to the condensing surface with the notches near the sides of the surface was made to realize large heat flux with high precision of measurement. As a result of measurement, the condensation characteristic curves over a wide range of vapor velocity (0.3∼68 m/s) at three vapor concentration of ethanol (mass fractions of 0.09, 0.32 and 0.53) were revealed. Further, the effect of the departing drop diameter on the maximum heat transfer coefficient, which was the representative point of the heat transfer of Marangoni condensation, was clarified quantitatively.

View full abstract

A numerical study is made on three-dimensional unsteady natural convection in a porous medium within a horizontal annular layer. The inner cylinder is maintained at constant and uniform temperature, and the outer cylinder is maintained at constant temperature with gradient β in the direction of cylinder axis. It is found that thermal oscillation for a low Ra_sDa_s range (Ra_sDa_s< 200) are caused at the upper part of the annulus due to the horizontal temperature gradient, which produces the traveling of the secondary convective cells.

View full abstract

Characteristics of the temperature field in forced convection turbulent heat transfer in a square duct with one rib-roughened wall have been investigated experimentally. Distributions of the mean temperature, primary flow velocity, intensities of the fluctuating temperature and velocity, correlation coefficients between the fluctuating temperature and velocity components, turbulent heat fluxes and turbulent shear stresses were measured, and they were compared with those obtained in a smooth-walled square duct to examine the influence of the rough wall on the characteristics of the temperature field. Near the rough wall, a weak dissimilarity appeared between the mean temperature field and the mean velocity field ; this suggested that the transport of heat by turbulence becomes less active than that of the momentum there. Based on the measured results, the mechanism of this dissimilarity in the mean fields was examined by evaluating the production terms in the transport equations of <u₁²>^^- and <t²>^^-.

View full abstract

Effects of shape of bluff body on suppression of turbulent Karman vortex street due to positive buoyancy are elucidated experimentally as follows : Almost of the turbulent energy in an isothermal wake is maintained in a positively buoyant wake. The vortex vorticity in a turbulent wake behind a prism cylinder decreases more remarkably due to positive buoyancy than a circular cylinder wake. In a turbulent wake, the critical Froude number, Fr_crit, where the Karman vortex street is suppressed due to positive buoyancy, of a prism cylinder wake is the same as that of a circular cylinder wake. Shape of a bluff body therefore has no effect on Fr_crit in a turbulent wake. Furthermore, the Karman vortex street in a turbulent wake behind a prism cylinder is suppressed more easily than a laminar wake.

View full abstract

Analysis has been made for the pulsatile flow of gas and the accompanying heat transfer within a regenerator made from mesh screens. A flow model is developed taking the mesh as a non-Darcy, thermally non-equilibrium porous medium. A harmonic analysis technique is used for solving the fully developed but unsteady gas flow in the regenerator. Based on the flow solution thus obtained, slowly evolving axisymmetric unsteady thermal fields are solved numerically over wide range of frequency by making use of finite volume method. Presentation is made of the friction factor and regenerator effectiveness. Effects of Reynolds number and frequency on the temperature profile are discussed. Importance of a thermal time constant of the system and length-to-radius ratio of the regenerator are demonstrated.

View full abstract

Void fraction measurement by high-frame-rate neutron radiography and the error estimation were conducted. Measurement error on instantaneous void fraction and time-averaged void fraction of forced-convective subcooled boiling was evaluated experimentally and analytically. Measurement error would be within 18% and 2% for instantaneous void fraction (measurement time is 0.89 ms), and time-averaged void fraction, respectively. Void fraction distribution of subcooled-boiling was measured using atmospheric-pressure water in rectangular channels with channel width 30 mm, heated length 100 mm, channel gap 3 mm and 5 mm, inlet water subcooling from 10 to 30 K and mass velocity ranging from 240 to 2000 kg/(m²·s) . One-side of the channel was heated homogeneously. Instantaneous void fraction and time-averaged void fraction distribution were measured par-

View full abstract

The mutual diffusion coefficients of H₂ into air and its component gases (N₂ and O₂) have been measured in the temperature range 30∼180°C and at atmospheric pressure by the use of the Taylor dispersion method. For the H₂-N₂ system, the composition dependence of the mutual diffusion coefficient has been studied by using both H₂ and N₂ as the carrier gas. The mutual diffusion coefficients for the H₂-air and H₂-N₂ systems are almost the same and smaller than that for the H₂-O₂ system by about 5%. The present data for the H₂-N₂ system, corrected to equimolar composition, agree well with the accurate data by the groups of P. J. Dunlop and of R. J. J. Van Heijningen.

View full abstract

The objective of this study was to fabricate a self-humidifying fuel cell stack humidified with water recovered at the cathodes, composed of 10 cells with 104 cm² cell areas, measure and simulate the performance of the stack. The model for the simulation is a three-dimensional model of the heat and mass transfer of water and gaseous reactants in fuel cell components with water cooling. The results of the stack experiments showed a maximum power of 250 W at the current density of 0.5 A/ cm². The simulation model showed good agreement with the actual performance of the stack. A self-humidifying stack with a vapor permeating membrane showed a performance comparable to conventional stacks and it is very effective in simplifying stack systems. The numerical analysis showed that the stack performance is affected by the anode and cathode gas flows, co-flow is superior to cross-flow and that one cooing cell is necessary for two or three generating cells to maintain a fuel cell temperature below 100°C.

View full abstract

In this experimental study, measurements were made on the characteristics of evaporating steady-state fuel spray injected into a high pressure and temperature cylindrical vessel by means of an electronically controlled common rail injection system. A fuel consisting of two components, n-octane and n-hexane, having different boiling points was used to investigate the characteristics of evaporation and mixing with the surrounding gas. Fluorescence characteristics of dopants excited by the YAG laser were calibrated by means of a spectroscope having an optical fiber. From the measurement of fluorescence wavelength, we decided that fluorobenzene and diethy-methyl-amine (DEMA) in a solution of fuel. Then, two-dimensional images were analyzed for each fuel component at liquid and gas phases, simultaneously. The results show that the density and temperature of the surrounding gas are two dominant factors influencing the penetration and evaporation of the spray.

View full abstract

Validity of numerical simulation for pulverized coal combustion is assessed, and effects of moisture in coal on the pulverlized coal combustion are numerically investigated. The results show that the combustion characteristics for bituminous coals are in general agreement with the measurements provided that moisture content in the coal is zero. With increasing the moisture, flame temperature and O₂ and NO_x mole fractions in the region near the burner decreases, increases, and decreases respectively, and the positions where the flame temperature and NO_x mole fraction show the maxima shift downward. Also, the increase of the moisture causes the increase of unburned fraction and the decrease of conversion ratio to NO_x. It is found that the contribution of thermal NO_x to the total NO_x drastically decreases with increasing the moisture, whereas the emission of fuel NO_x increases.

View full abstract

Using a high-speed video camera with an image intensifier, observations have been made on the propagating flames in the vertex rings. From their intensified images, variations of the flame speed and the flame diameter along the vortex axis have been determined. Results show that there are four types of flame propagation, (1) steady flame propagation, (2) periodic flame propagation, (3) unsteady flame propagation, and (4) random flame propagation. The steady flame propagation occurs for the maximum tangential velocity V_θmax>7 m/s (Re<__-10⁴) and its local flame speed is almost constant during the propagation. The periodic flame propagation is originated from the longitudinal instability of the vortex ring, which occurs around V_θmax &ap; 7 m/s in this experiment. (3) The unsteady flame propagation is accompanied with strong flame acceleration and deceleration, while its mean flame speed from ignition to the end is almost equal to the maximum tangential velocity. The random flame propagation occurs in turbulent vortex rings. The flame speed changes randomly. The luminous zone is quenched shortly behind the flame and its mean flame speed becomes much lower than the maximum tangential velocity.

View full abstract

A hybrid probability density function (PDF) method has been developed so as to analyze turbulent diffusion flames, devoting to only PDF's of the mixture fraction, based on the conserved scalar approach. To evaluate the potentiality of the method, a turbulent jet diffusion flame has been calculated by this method and also the presumed PDF method called the flamelet model to compare those results with experimental results. The hybrid PDF method is able to evaluate the process of inhomogeneous turbulent mixing, while the flamelet model overestimates the scalar turbulent mixing, as the presumed PDF assumption ignoring the inhomogeneous mixing. The accuracy of the estimation of the flow field is crucial for the hybrid PDF method, whereas the underestimation of the turbulent kinetic energy by the k-ε 2 equation model used in the present study leads to the underestimation of the scalar turbulent mixing.

View full abstract

Transition processes from the edge flame without premixed flame to the edge flame with premixed flame, the triple flame, of methane-air edge flame are experimentally investigated by using the novel burner system. The burner system forms a diffusion flame in an axisymmetric wall jet. The measurements of the stability limits and the flame locations, and the observation of the flame shapes have been performed. The burner can make three types of the stable edge flame. First one is the edge flame without premixed flame, whose location is determined by the wall interaction with the edge flame, that is, the thermal factor. Second one is the edge flame with premixed flame, whose location is still determined by the thermal factor. Third one is the edge flame with premixed flame, and the location is determined by a balance between the propagation rate of the edge flame and the local velocity of unburnt mixture, that is, the fluid dynamical factor. The change of the dominant factor for flame location is due to the increase in the partially premixing region ahead of the edge flame. The increase in the partially premixing region makes it possible to form premixed flame and the first edge flame changes into the second one. Further increase in the partially premixing region lead to an increase in a total amount of the heat release around the edge flame region and the flow field ahead of the edge flame diverges. Finally, the edge flame is lifted up by the flow divergence and the dominant factor to determine the edge flame location changes from the thermal factor to the fluid dynamical factor.

View full abstract

This paper reports PIV measurement of diesel fuel sprays injected from a single-hole nozzle at high-injection pressure comparable to operating conditions of commercial diesel engines. The injection pressure ranges from 30-70 MPa. The fuel is injected into a non-combusting atmosphere, which is charged up to 2.0 MPa. A laser-induced fluorescent technique is employed to visualize internal structures of fuel sprays formed by densely distributing droplets. A direct cross-correlation PIV technique is applied to the spray images in order to extract droplet velocity information. Unique large-scale structures, called 'branch-like structures', are observed and shown to be associated with active vortical motions of droplets, which appear to be responsible for the mixing between droplets and surrounding gas. Preferential concentration of droplets is found to occur by the centrifugal force exerted by the vortical motions.

View full abstract

An ignition and combustion model has been developed to predict the heat release rate in direct injection compression ignition engines employing very ealy injection timing. The model describes the formation of ignitable or combustible fuel-air mixture by turbulent mixing, and includes five chemical reactions, including low-temperature oxidation. The KIVA II computer code was modified with the present ignition and combustion model. The numerical results obtained by the modified code indicate that the model developed in this work reproduces major features of two-stage autoignition, as well as experimentally observed trends in NO_x and unburned fuel emissions. The computational results showed that NO_x emissions are significantly influenced by fuel injection timing. The results also indicated that fuel droplets which enter the squish region possibly become unburned fuel emissions. The relationships among the in-cylinder fuel spray distributions, fuel-air equivalence ratio, temperature, and mass fractions of NO and unburned fuel can be demonstrated by the graphical results included in this work.

View full abstract

Recently, pre-mixed lean mixture combustion has attracted attention, and many researchers are investigating compression ignition of a pre-mixed lean mixture. Previous research in our laboratory has shown that low NO_x emission combustion is obtained by PREDIC (PREmixed lean DIesel Combustion). However the operating region of PREDIC is limited to partial loads at low engine speed. So the operating region was researched in detail under the higher engine speed and the expansion of PREDIC operating region was tried. As the result, PREDIC was obtained at half load condition of all engine speed. In order to extend the operating region of PREDIC to the higer load conditions, avoiding the earlier ignition is necessary at high load. EGR, low compression ratio and supercharging are effective to increase the power output. And that the fuel consumption was depend on an excess air ratio, compression ratio, and boost pressure.

View full abstract

This paper investigates engine performance with an emulsified fuel including used frying oil composed of vegetable oils discarded from restaurants and households. A single cylinder DI diesel engine was operated with transesterified fuel from used frying oil, the so called "biodiesel". Compared with gas oil, the BSEC of neat biodiesel improved at high loads and retarded injection timings, while the smoke density was reduced at all operating conditions. The engine performance with fairly stable emulsions of biodiesel and water was also examined, and emulsified biodiesel with 30% (vol.) water showed a significant reduction in NO_x (1 100 to 400 ppm) while maintaining the minimum BSEC value achieved with gas oil. It was also found that using biodiesel emulsion at a rated output, the trade-off relation NO_x vs. BSEC and NO_x vs. smoke improved slightly over the gas oil emulsion.

View full abstract

Recently, the electronic-hydraulic control fuel injection system for diesel engines has potential as a means of reducing NO_x and PM simultaneously, and is already being used in small engines in Japan. However, it can not yet be used practically in the case of large engines because fuel consumption, electronic-hydraulic control power, and manufacturing costs must be reduced, and also improvements in reliability are required. Needless to say, the point of the system is an injector. We applied Pascal's principle to this study and made a new intensifier style injector which controls fuel injection in one solenoid. We examined those reduction measures by calculations and experiments with the new injector. Ultimately, it was proved that the further improvement of a two position, three port connection valve was a conclusive factor of the system.

View full abstract

The effect of nozzle orifice size on smoke emission was examined experimentally in a turbocharged DI diesel engine. The nozzle orifice size was varied from 0.26 to 0.18 mm under various engine operation condition. The rates of NO-formation, soot-formation and soot-oxidation were analyzed by using the two-zone model based on the measured combustion pressure time history. The changes due to orifice size in the excess air ratio of the burned zone, the soot particle size and the number of soot particles were estimated. It is shown that the soot size is primarily dependent on the orifice size and secondarily but more markedly on the spray penetration for the ignition delay period, and the number of soot particles is primarily dependent on the amount of fuel injected and secondarily on the excess air ratio of the burned zone.

View full abstract

The effect of triangular plates in a mixing channel on fuel-air mixing was investigated to design a compact gas-turbine combustor. Turbulent flow was visualized by laser-light sheet method and velocity distributions were measured by cross-correlation particle image velocimetry. Also, fuel distributions were measured by Rayleigh scattering method. The obtained results show that triangular plates produce a large-scale fluid motion perpendicular to the mean streamline together with small eddies. They might stir the bulk flow and promote the turbulent mixing downstream, thus resulting in a more homogeneous mixture, whereas in the case without triangular plates fuel distributes heterogeneously and the variance of mixture concentration is high even in the flow downstream.

View full abstract

The input to be supplied to Combined Cycle Waste to Energy (C·C-WTE) system is waste incineration heat and combustion heat of gas turbine fuel. While the maximum temperature of working fluid by the former is usually 300°C, the temperature of that by the latter is approximately 1 200∼1 300°C. That is, the energy importance which the two heat gives to working fluids is much different. Therefore, an evaluation of C·C-WTE system based on heat capacity without energy importance sometimes brings unreasonable result from thermo-dynamics viewpoint. In this paper, C·C-WTE systems have been evaluated on exergy to be received by water/steam in waste heat boiler by incineration gas and chemical exergy of fuel to be inputted to gas turbine. The differences between exergy efficiency to be considered their energy worth, heat capacity due to enthalpy and repowering efficiency which has been studied in the former paper have been compared.

View full abstract

Lamimar natural convection along the outer surface of a vertical cylinder, heated uniformly, was numerically investigated for varying transverse curvature of cylinder and Prandtl number of fluid. Local and average Nusselt numbers at Prandtl number of 0.02, 0.7, 7.0, 100 were calculated for the curvature of cylinder r₀/L (ratio of radius to length) 0.005, 0.012, 0.025, 0.050, 0.500 at the modified Grashof number Gr_L^* from 5×10⁷ to 5×10¹⁰. A study was also conducted experimentally for air of Prandtl number of 0.7. The numerical values of local Nusselt number Nu_x are compared with existing analytical and numerical results as well as experimental ones.

View full abstract