A linear analysis based on a semi-actuator disk method is carried out to investigate the effects of flow compressibility due to 2-phase flow on rotating cavitation. It is assumed that no slip and no phase change occur between gas and liquid phases, and that the flow field is isothermal because of the large thermal capacity and conductivity of the liquid phase. The results show that there are many solutions of characteristic equation, corresponding to conventional forward and backward rotating cavitations, rotating stall, solution with no disturbances, natural vibrations of 1-D cascade flow resonant with the upstream and downstream flow fields, and a forward rotating mode different from conventional forward rotating cavitation. It is also shown that the stable regions of conventional rotating cavitations become slightly larger due to the flow compressibility.

View full abstract

Lattice vortex method by introducing virtual vortices has been proposed in this paper. In our method, a staggered lattice for the velocity and the vorticity is used and the virtual vortices are placed to approximate the velocity field obtained by the Biot-Savart law. Velocity field can be determined by calculating only the induced velocities from the neighboring vortices. This point is completely different from the conventional method using the Biot-Savart law and the calculation time was almost half compared to that of using the Biot-Savart law. The discretization error of our method is proved to be the second-order by taking an adequate distribution of the vortices and the result of calculations of Burgers' equation agrees well with the theoretical solution. The calculation result of a cavity flow shows a good agreement with the result obtained by the conventional finite difference method. The lattice vortex method is shown to be useful for not only outer flows but also inner flows.

View full abstract

This study investigated the large vortex shedding mechanism behind a wedge at three angles of attack (α=0°, 22.5°and 45°) having the different separation patterns. This flow visualization of the velocity and the vorticity was executed by the LDV measurement and the FFT-spectrum processing technique. The vortex attained full growth at the end of the vortex formation region. A stagnation point existed on the boundary of the region and almost coincided with the point, where the large vortex intersected the separated shear layer. Specially at α=22.5°, a separation bubble formed and stayed on the upper side of the wedge. As compared with α=0°and 45°, the bubble made the vortex formation region longer and flatter. In addition, the universal wake Strouhal number defined by Griffin was larger. Fluctuating velocities with high frequencies were generated on the separated shear layer surrounding the bubble.

View full abstract

Flow field around a bathtub vortex was investigated. A finite cylindrical vessel was used as a test section. The vessel had a tangential inlet open channel at the top, and a vertical outlet pipe at the bottom center. Flow field was visualized by ink dye and particle tracers, resulting in velocity distributions to be measured. The flow inside the vessel was mainly divided into two parts ; one was through the central region of the vessel, and another was along the wall of vessel, i.e., the side and bottom walls. Flow through the central region was found to increase its velocity in proportion to a distance from the free surface and to form a stagnant region along the vortex center under certain experimental conditions. The flow field and the vortex profile were determined by strength of circulation and flow rate in the central region. However, these local parameters complicatedly depended on the macro parameters such as total flow rate, the outlet pipe diameter and the depth of the vessel.

View full abstract

Three dimensional flow field around a circular cylinder was analyzed by using a third-order upwind finite-difference method for Reynolds number=1000. Lift and drag force spectrum have some peaks near main karman frequency. It was also found that streamwise velocity and pressure spectrum around separation point have some peaks with a difference mode of these peaks (beat mode). Spanwise flow structure is formed by such modes that have nearly half frequency of Karman mode (half mode) and which are created by three-dimensional vortices. In the present analysis, three dimensional vortices induce half modes and a beat mode additional to Karman mode, resulting in the observation of complex three dimensional flow field.

View full abstract

Numerical simulations have been carried out for a two-dimensional falling liquid film with interfacial waves. The staggered grid fixed on a physical space is employed and basic equations, continuity and Navier-Stokes equations, as well as equations for boundary conditions are discretized on the staggered grid without omitting any term in the equations. Neglecting the interfacial shear stress is only an assumption for the present analysis. New numerical methods are proposed for the interfacial boundary conditions and for the change of the film thickness, derived from the continuity equation. Simulation results agree well with experimental results in which waves are formed by artificial periodic disturbances. For both the simulation and the experiment, the developed waves disturbed at low frequency have a structure combined of a roll wave and capillary waves, while there is no capillary wave for high-frequency disturbance. The adyantages of the proposed methods are also presented.

View full abstract

To apply the direct numerical simulation (DNS) and the large-eddy simulation (LES) of turbulence to flow fields of complicated geometry, a higher-order finite-difference method (FDM) has been developed for the body-fitted coordinate system. The consistency and the conservation property of FDMs are discussed for the collocated grid. As numerical examples, DNS results of decaying isotropic turbulence and DNS/LES results for plane channel flow are shown and the influence of variable arrangement is examined. The results by the consistent 'interpolation' method for gradient form on the collocated grid agree well with those by other proper FDMs and the spectral method.

View full abstract

Incompressible viscous flows around a three-dimensional ring at Re=50 where the ring axis is parallel and vertical to the free stream have been numerically investigated. A Navier-Stokes equation solver employing an overset grid system was applied to treat the complex geometry. The following things were elucidated. When the ring axis is parallel to the free stream, the pressure on the ring axis obeys a similarity relation. The minimum pressure on the axis occurs just beyond the throat of the ring. The stagnation point exists in the inner side of the ring. The aerodynamic force in the ring plane directs outwards. This force and the drag coefficient decrease as to a smaller ring radius. When the ring axis is vertical to the free stream, the flow patterns were clarified for both the stationary and the rotating cases. To compare the two cases, the stagnation point moves in the direction opposite to the rotation. The stagnation pressure and the pressure in the rear part of a rotating ring are lower than in the stationary case due to the centrifugal force. Lift occurs in the rotating ring and the drag coefficient is slightly larger than that in the stationary case. Both lift and drag coefficients are smaller than in the two dimensional case.

View full abstract

In this paper, Free Mesh Method (a kind of meshless method) proposed by the authors is improved for application to steady incompressible viscous flow analysis. In the finite-element analysis of incompressible viscous flow, it is better to use different-order interpolation function for velocity and pressure. In the method, quadratic interpolation function can be used by introducing medium nodes between two original nodes, and a set of advection, diffusion and gradient matrices can be assembled independently node-by-node. As a numerical example, a two-dimensional driven cavity flow is analyzed. In this example, the governing equations, i.e. a coupled system of incompressibility and Navier-Stokes equations, are solved simultaneously. To solve a set of linear equations derived from the formulation of this method, the conjugate residual (CR) method is used. It is concluded that the result obtained using the method agrees well with the numerical result obtained by Ghia et al.

View full abstract

Detailed flow structure of a deep rotating stall cell in a single-stage axial compressor is studied based on a numerical simulation of simplified quasi-three-dimensional compressible Navier-Stokes equations a linear cascades system and unsteady flow measurements in a test compressor. The computational system is expanded to a maximum combination of 15 rotor blades/25 stator vanes, so that cell growth is not hampered by spatial constraint of the field. The simulation is performed at a blade Mach number of 0.3 to meet the rig test condition. It is found that the stall cell is composed of sub-cell vortices, each extending over 3-5 rotor pitches, which align circumferentially ahead of the rotor, and counter-rotating vortices occupying the stator passages. The computed cell speed agrees well with the measured speed. In the computed time trace of flow velocity at the rotor inlet, multiple peaks, corresponding to the sub-cell vortices, appear during a cell passage. These are also seen in the measured traces taken by split-film probes. The agreement indicates that multiple sub-cell vortices may constitute a deep stall cell in the real situation.

View full abstract

This paper presents the detailed experimental and numerical analyses for the flow characteristics of an annular air jet and its control using a coaxial round jet. As a result, the mean and fluctuating flow characteristics of an annular jet, such as the velocity profiles, decay of maximum velocity, spread of jet, flow rate, pressure distribution, turbulence intensity and turbulence kinetic energy, and the effects of a coaxial round jet on them were clarified, and it was found the flow characteristics of the annular jet can be controlled by changing the velocity ratio of the coaxial round jet to an annular one. Numerical analysis was carried out using the k-ε turbulent model.

View full abstract

This paper presents a numerical procedure to predict sand erosion and the interaction between the flowfield and the eroded surface. To simulate this phenomenon, the turbulent flowfield, the particle trajectory and the amount removed from the eroded wall are calculated iteratively. In computations of the flowfield, incompressible Reynolds equations and a high-Reynolds-number type k-ε turbulence model are adopted. Assuming that the particle suspension is dilute, the collision of particles and the influence of particle motions to the flowfield are ignored. The erosion model proposed by Neilson and Gilchrist is used to calculate the weight loss. To verify this method, a 90°curved channel and a channel with a bump are computed. The results show that the present procedure can reproduce the sand erosion process and the temporal change of the flowfield qualitatively.

View full abstract

When a Shinkansen train passes through a tunnel, the pressure change due to the pressure wave in the tunnel penetrates the cars. This gives rise to aural discomfort of passengers, which becomes severe when trains pass each other in the tunnel. In this paper, a new ventilation system for alleviating this aural discomfort is described. A method of numerical simulation of the pressure change outside and inside of the train is also presented. A field test to confirm the fundamental performance of the system has been carried out using Shinkansen electric cars Series 300 and the results are compared with those of the numerical simulation. Moreover, the applicability of the system to a super-high-speed train travelling at 350 km/h is ascertained by numerical simulation.

View full abstract

The final goal of this research is to establish fundamental technologies for active reduction of fan tones radiated from a turbofan engine. The acoustic system developed comprises a modal detection system, a high-speed signal processing system and a secondary sound source system. The secondary sound source system, which is required for anti-phase acoustic mode generation, is the main topic of this paper. We designed the system to generate and identify arbitrary spinning modes in a duct and examined the performance of the system by conducting experiments. As a result, the system generated spinning modes within the range of 1∼7 lobes. A flush-mounted microphone array could analyze radial modes of higher order as well as circumferential ones with good accuracy. It was clarified that the system can be used as a secondary sound source for the active suppression of fan tones.

View full abstract

Using an energy conversion device with a magnetic fluid susceptible to temperature in a nonuniform magnetic field, the effects of particle aggregation in the magnetic fluid on the flow characteristics and effciency of the device are investigated experimentally and theoretically. The experimental data showing the effects of the strength and direction of the magnetic field and temperature difference on pressure difference and efficiency are compared. These experimental data are compared with the theoretical results by taking into account the particle aggregation, assuming one-dimensional flow. By estimating not only the magnetic pressure but also the hydrodynamic drag deu to the increment of apparent viscosity by applying the magnetic fields to the device, the effects of the aggregation on the flow characteristics and efficiency are elucidated.

View full abstract

Supersonic combustion using a catalytic combustion in a cold supersonic flow field was investigated in a square duct with a backward-facing step. The free stream Mech number was M_m=1.81. Hydrogen was injected transversely behind a backward-facing step into a cold air free stream. Using a catalyst in a cold supersonic turbulent mixing layer, it was found that hydrogen reacted stably to oxygen in the air flow. The relationship between the heat release due to catalytic combustion and supersonic flow properties, which influence the supersonic combustion, was clarified experimentally. The spatial distribution of heat release generated by catalytic combustion in the supersonic turbulent mixing layer is discussed. It was found that the heat release due to the catalytic combustion had a maximum at the outer edge of the mixing layer.

View full abstract

In this paper, the influence of shape change of the impeller and scroll in the axial direction on performance and noise are investigated. It is shown that a 14% increase in the hub-side diameter of the impeller over that of the shroud side leads to a 4% increase in the efficiency η_i and a 2dB reduction in the specific sound level SLs. It is also shown that an increase in the diffusion angle of the hub side of the scroll to that of the shroud side raises the efficiency by 2%.

View full abstract

The effect of an unsteady hydraulic force on an impeller of a mixed-flow pump with a double volute casing that has an unsteady pressure distribution is experimentally investigated to show the relationship between pressure fluctuation and unsteady radial and axial thrust. The results show that (1) the time averaged axial thrust is caused primarily by the total head of the pump and secondarily by the effect of the flow at gaps between the impeller and the casing, (2) the unsteady radial and axial thrust is induced by the change of instantaneous pressure distribution in the volute casing, (3) the pattern of this pressure distribution is characterized by the operating flow rate of the pump, and (4) the fluctuation amplitude of the radial thrust depends on the number of blades or the pump configuration, whereas, that of the axial thrust is independent of these factors.

View full abstract

Spectrum of the plume in a stably stratified ambient was elucidated to have gradients -9/2 and -8.0 at laminar state, and -5/3 and -3.0 at turbulent state, in addition the frequency band of the turbulence was higher than that of the swaying motion. By employing the above results, flow regimes, i.e. laminar, transitional, and turbulent, at any location can be determined. By plotting flow regimes on visualization photos, retransitional and relaminar phenomena are specified. Stable stratification generates turbulence, suppresses turbulence, and leads to reverse transition and relaminarization, which do not occur in the plume in an unstratified ambient, thus characterizing the plume in a stratified ambient. The Grashof number for the beginning of transition is approximated as Gr_s=9.15×10⁵{Q[W/m]}¹.62. Critical heat rate Q for the beginning of either the transitional or turbulent regime is Qs[W/m]=1.30-2.75 or Qe[W/m]=5.10-8.73.

View full abstract

A sensitivity analysis for general evaluation of the influence of thermal boundary conditions on the natural convection heat transfer is proposed. In order to seek the sensitivity function, we employ the perturbation formulas for the governing equations of the natural convection field and derive their adjoint operators. Then, the sensitivity function is obtained using the numerical solutions of the unperturbed heat transfer problem and the perturbed adjoint problem. As a result, the change of total heat transfer from an original one can be calculated by an inner product of the sensitivity function and the perturbation of the thermal boundary conditions. The sensitivity functions for the natural convection heat transfer in a square cavity are presented to demonstrate the proposed method.

View full abstract

The effects of an ultrasonic wave on nucleate-boiling heat transfer, focusing on depression of the incipient boiling temperature were examined. Experiments were conducted using a copper thin film and saturated R-113 liquid for a pool condition at 0.10 MPa. The incipient boiling temperature was depressed by ultrasonic wave incidence up to 10 K in reheating experiments where the heat transfer surface had been immersed in the liquid following the previous boiling experiment. On the other hand, it was minimally affected when the boiling experiment started immediately after the test surface was immersed into the liquid. These results were considered to be related to the number of active nucleation sites available. The decrease of the incipient boiling temperature as the power of the ultrasonic wave was increased, however, did not depend on the frequency. It was pointed out that the depression of the incipient boiling temperature was caused by the local pressure increase caused by the ultrasonic wave incidence.

View full abstract

Two dimensional calculation was performed for combined convection heat transfer in a channel with two ribs attached to one wall, following the previous study on forced convection case without buoyancy. The flow is heated from the surfaces of both ribs and the present study dealt with the two cases of buoyancy-assisting flow and buoyancy-opposing flow. The effect of Reynolds number, Re_L, and modified Richardson number, Ri^*, was examined keeping space between ribs, σ, and blockage ratio, τ, constant (σ=3.0, τ=0.5). Increasing the magnitude of buoyancy, unsteady flows predicted by the present calculation are stabilized in both of two cases. Serious deterioration of Nusselt number on the 2nd rib suddenly occurrs in a certain range of Ri^* due to the flow stabilization. This is because flow unsteadiness plays an important roll for heat transfer enhancement as was described in the previous study. However, in buoyancy-assisting flow, similar deterioration of Nusselt number also appears on the 2nd rib even if flow remains steady. This is caused by the disappearance of strong rotating flow which exists in the cavity between the both ribs and keeps fluid in the cavity cooler.

View full abstract

Optical interferometers, such as the Mach-Zehnder interferometer, have often been used to measure the temperature of transparent fluids. In this paper, we utilize a Moire interferometer in which a conventional interferometer is combined with the Moire technique. By applying a fringe scanning technique to this Moire interferometer, a new fringe analytical system is obtained and the fringe analysis is automated with the aid of a personal computer. The temperature field of the air around a vertical heated plate under natural convection is measured and compared with the theoretical solution. The temperature measured using this system is in good agreement with the theoretical one. It is verified that this optical system is applicable for measuring the temperature of transparent fluids.

View full abstract

Ice formation phenomena as well as freeze-shut conditions of flowing water in an isothermally cooled parallel plate channel are examined. An ethylene glycol 4.6 wt% solution is also used as a test fluid to compare with the results of water. It is found that the ice formed in the solution flow is different from that in water flow in thermal properties, and that the mushy ice formed in the solution at the expanded flow regions affects significantly the freeze-shut conditions. The present analytical results for freeze-shut conditions of water as well as solution flows agree with the experimental data, and show that an ethylene glycol 4.6 wt% solution is a little harder than water to make the flow freeze shut.

View full abstract

This paper deals with the experimental investigation of latent thermal energy storage characteristics of surface cross-linked form-stabilized high-density polyethylene pellets as a phase change material (PCM), using pool boiling and condensation of an ethylene glycol/water solution. The polyethylene pellets and heat transfer medium of ethylene glycol were mixed and poured into a cylindrical heat storage vessel. The heat transfer into the polyethylene pellets was mainly enhanced by the condensation of vapor bubbles of ethylene glycol/water solution which was heated by an electric heater set at the bottom of the vessel. The effects of heat input, the amount of PCM and ethylene glycol/water solution, and heater surface temperature on the heat storage characteristics were investigated. As a result, the nondimensional correlation equation of the completion time of the heat storage process was derived as a function of the nondimensional parameters.

View full abstract

Snow has been already used effectively as the coolant for refrigerator in some heavy snowfall area in Japan. As the surface of snow is covered with cold melting water, we can expect the gas absorption on the surface. By this mechanism, the snow is able to remove some contaminants from air such as dust or harmful gases. As in the snow cooling system, air is cooled directly by the surface of vertical snow hole drilled through snow pile, some contaminants are removed automatically and simultaneously with heat transfer process. The heat and mass transfer process are very complicated because the surface area of snow decreases and the shape of snow changes in accordance with its melting process. In this study, the authors measured the absorbability of ammonia gas as the example of contaminant of room air with a pilot plant of the snow cooling system. Experimental results show that the mass transfer rate of ammonia through the melting snow surface is depndent on air flow rate and ammonia concentration, but is not dependent on the volume of snow remaining. And it is cleared that this filter effect of the snow surface can be applied to actual use of contamination control of air conditioning.

View full abstract

An experimental study was performed to explore the characteristics of the leading flame edge of a flame spreading over a combustible material. The movement of the gasifying region of the combustible solid beneath the flame was simulated using a velocity comtrollable apparatus on which a porous plate ejecting combustible gas was installed. Methane, propane, and a methane/air mixture were used as the combustible gases. The characteristics of the leading flame edge were examined by changing the velocity of the apparatus from 0 to 100 cm/s after a stable flame was established over the porous plate. Furthermore, flowfields near the leading flame edges were examined using a particle tracer technique. It was found that the leading flame edge remained stable when the velocity of the apparatus, i.e., the velocity of the leading flame edge, was below a limiting velocity, between 20-30 cm/s in the present experiments (methane and propane). However the behavior of the leading flame edge of the premixed flame (methane/air) was different from at of the methane or propane flame. Although the position of the leading flame edge changed with the velocity, the leading flame edge remained stable to 100 cm/s.

View full abstract

The combustion characteristics of hydrogen jet diffusion flames on a flat plate in a cross flow were studied experimentally and two distinct types of hydrogen jet diffusion flames were observed. The first type was the hole opening flame, which had a hole on the front side of the flame at the transition point from laminar to turbulent flow. It was found that the turbulence intensity was suddenly increased at the lowest position of the hole. The hole opening phenomenon may be explained by the mass and heat transfer rate exceeding the finite chemical reaction rate of hydrogen with air. As the hydrogen injection velocity increased, the hole opening flame changed to the second type of flame. This was a stick flame which formed along the recirculation zone of the hydrogen jet acting on the cross flow as a "rod". Detailed observations showed that a premixture was formed in a mixing layer between the hydrogen jet and the surrounding cross air flow, and it burned just behind the hydrogen jet.

View full abstract

In this study, a new ceramic-opposed jet burner was designed to ensure the adiabaticity of the flame. In the opposed jet turbulent premixed flame, the concentrations of stable species and temperature distributions and their structures have been investigated. The opposed jet burner flame can persist far beyond the stretch rate, at which point the wrinkled laminar flame is extinguished. The results show that the flame structures consist of stretched stringy vortices involving various reacting species (e.g., C₃H₃, hydrocarbonsns, O₂) which are referred to as the distributed reaction zones, and that the flames are divided into two categories : (1) transition flame from a wrinkled laminar flame to a distributed reaction zone, and (2) a fully developed distributed reaction zone. The structure of the distributed reaction zone depends mainly on the mixture supply velocities and is insensitive to the equivalence ratios. The stretched stringy vortices of the distributed reaction zone eradicates the nonuniformity of temperature and concentration fields. The hydrocarbon species as well as fuel are consumed completely in the distributed reaction zone, and almost complete combustion is achieved.

View full abstract

A numerical procedure to predict acoustic characteristics of a two-dimensional valveless pulse combustion system is described in detail, introducing a plane sound wawe approximation. The resonance sound pressure in the combustion chamber is numerically estimated by applying plane sound pressure with various frequencies in the combustion chamber and by synthesizing reflected sounds from both the upstream supply line and downstream exhaust line under the cold and stagnant conditions. In this paper numerical analysis is first performed on a simple Helmholtz-type combustor consisting of a combustion chamber and a tailpipe. Experimental measurements of acoustic characteristics of the combustor are also made. By integrating suitable values for the acoustic parameters, such as the damping factor and reflection coefficient, into the calculation, fine agreement of the numerically predicted resonance characteristics with the experimental measurements is obtained, indicating the usefulness of the proposed numerical approach for examining acoustic characteristics of the practical pulse combustion systems.

View full abstract

A study was carried out to determine the relationship between sound pressure generated by open premixed flames and the velocity fluctuations in an unburnt mixture flow at a cylindrical single-port burner. The velocity fluctuations were measured at the burner port. The turbulent energy flux Ek. using the measured velocity fluctuations, was defined as an index of the velocity fluctuations. It was found that sound pressure generated by turbulent premixed flames on the burner can be determined by the turbulent energy flux Ek at the burner port, and that it is proportional to Ek^*. The maximum sound pressure was generated by turbulent flames having an air ratio between 0.8 and 0.9. The effect of the laminar burning velocity on sound pressure was also investigated by changing the laminar burning velocity. It was found that the higher the laminar burning velocity of the mixture, the more sound pressure was generated from the turbulent flame for the same turbulent energy flux Ek.

View full abstract

A mechanism of changing the local burning velocity of a stretched premixed flame independent of change in the local heat release rate per unit flame surface area is clarified. Based on the mechanism, the dependence of the local burning velocity on the local fuel consumption rate and the local flame stretch is derived mathematically for a stationary flame of the Lewis number of unity, by considering the mass ballance of the fuel entering and consumed in a stream tube which crosses the flame surface. Unlike the equations theoretically derived so far, the proposed equation is valid without the limitation of small stretch rate and small flame curvature, and therefore it can predict the local burning velocity of the stretched flame of the previous numerical study (Waku et al., 1997) very well for a wide range of flame stretch.

View full abstract

A series of experiments about the ignition of H₂-O₂-O₃/H₂-O₂-O₃-Ar mixtures with KrF excimer laser has been carried out. We invesitigated the influences of some factors upon the Minimum Incident Laser Energy Intensity (MILEI) for the ignition, and took the schlieren photographs with a high speed camera to observe the photolysis and the ignition process. The results of these works showed that ozone was only a considerable absorber for the focused laser light and suggested that the photolysis of ozone induced the ignition. In our experimental range, the MILEI decreased with the increase of concentration of ozone, dilution with argon and preheat of mixture flow. However, a weak dependence of MILEI on the equivalence ratio of H₂-O₂-O₃ mixtures was measured. From the schlieren photographs, we observed the rapid growth of initial ignition kernel and the formation of shock wave. These observations suggested that the ignition process was quite faster than the one induced by photolysis of O₂ with ArF laser.

View full abstract

The purpose of this investigation is to clarify the atomization mechanism of a high speed liquid jet which issued from a single hole nozzle. The breakup process of the liquid jet is thought to be greatly affected by the behavior of the internal flow in the nozzle hole. In order to investigate this, visualization of the flow in the nozzle hole, observation of the atomization of the liquid jet and measurement of the static pressure in the nozzle hole were performed under atmospheric conditions using transparent nozzles with various length-to-diameter ratios of holes, L/D. A non-dimensional pressure coefficient in the nozzle hole was defined by C_p=(p_c-p_v)/(p_a-p_c), where p_c is the static pressure at 1mm from the inlet of the nozzle hole, p_v is vapor pressure of water at 293 K and p_a is ambient pressure. The behavior of the internal flow in the nozzle hole where cavitation occurs was arranged by using c_p. It was clarified that the primary factor in the atomization of the liquid jet is the disturbance of the liquid flow resulting from cavitation phenomena.

View full abstract

To reduce exhaust emissions in high-speed DI diesel engines, it was reported that a high swirl ratio is required for partial load operating conditions. In this paper we discuss the high swirl helical intake port necessary to put this combustion concept into reality. The impulse swirl method, instead of the paddle wheel (vane) method, was determined to be appropriate for estimating the high swirl ratio. To select an effective shape for the port, a parts-assembly port model was designed, and a parametric study of the helical port shape was carried out. An auto mesh generator for helical port numerical analysis was developed, and a 3-dimensional simulation of the intake flow pattern necessary to make a high swirl ratio was conducted.

View full abstract

Although the time-space mechanism of NO_x formation in diesel injection combustion at high temperature and pressure has conventionally been pursued by the sampling method, and so forth^[1], the cause-effect relationship of its formation factors remains to be more accurately clarified. In our present study, concentrations of NO, OH and soot in diesel spray flames from a single injection nozzle at high temperature and pressure were measured by use of a motoring compressed injection test system with application of laser-induced fluorescence^{[2], [3]} and laser-induced incandescence^[4], both of which can provide unperturbed measurements necessary for clarifying formation characteristics of harmful exhaust substances in diesel fuel flames at high pressure.

View full abstract

ATAC (Active Thermo-Atmosphere Combustion) is stable in the lean limit, because it is "bulk-like" and/or "nonpropagating" combustion caused by self-ignition. In a low-heat-rejection methanol-fueled engine, we investigated the influence of ATAC operation with weakening premixed gas and raising temperature of the combustion chamber wall on cyclic variation of ignition and combustion. We analyzed the cyclic variation of autoignition timing, combustion duration, combustion quantity and instantaneous heat flux on the wall of the combustion chamber, from which we clarified the correlation between autoignition timing and combustion duration, autoignition timing and combustion quantity, and combustion duration and combustion quantity.

View full abstract

In our previous work, we found that the preferential diffusion in a turbulent flame played an important role in its turbulent combustion characteristics, and estimated the local burning velocity in premixed turbulent combustion experimentally, taking account of the preferential diffusion effect. In this study, a model which takes the preferential diffusion effect into consideration, is proposed to predict the premixed turbulent burning velocity, using the local burning velocity as a reference instead of the original laminar burning velocity. The model can be explained as follows. The turbulence affects the turbulent burning velocity by increasing the flame surface area and stretching the flame. Consequently, the turbulent burning velocity and quenching limit are determined by the balancing of both effects. The predicted velocities are compared with the measured turbulent burning velocities such that the fuel, equivalence ratio and laminar burning velocity were varied extensively. As a result, quantitative accuracy of this simple model is confirmed.

View full abstract

In order to reduce the fuel consumption at partial load in an indirect injection (IDI) diesel engine, a new engine system was installed in a four-cylinder diesel engine, and tests were conducted to investigate its potential. It was clarified that enlargement of the jet passage area under the new combustion chamber with a refined jet passage to reduce heat loss ; use of higher temperature cooling water and lubricating oil ; and delay of the closing timing of intake valves in order to reduce friction loss are effective methods for reducing the fuel consumption at partial load. This new engine system was proven to reduce fuel consumption at partial load by approximately 20%.

View full abstract