This paper describes various technologies for compliance with diesel engine emission regulations up to the year 2000. It is anticipated that future diesel emission standards will be met by a combination of systems, including such innovations as high-pressure injection, injection rate control, strong turbulence in the later stages of combustion, cooled EGR (Exhaust Gas Recirculation), SCR (Selective Catalytic Reduction) and particulate trap filters. For further reduction of NO_x, the application of water may be required.

View full abstract

Accurate estimation of the area-averaged volumetric fraction is required when we evaluate two-phase flow characteristics such as pressure drop and heat transfer. A two-phase flow model based on local relative velocity was proposed in this study. The effects of distribution of local phase velocities and local volumetric fractions on the area-averaged volumetric fractions were also taken into account in the model by introducing new distribution parameters. In order to examine the applicability of the model to the estimation of the area-averaged volumetric fractions, measured area-averaged volumetric fractions of solid-liquid flow and gas-liquid flow in vertical pipes were correlated based on the basic equation of the model. The area-averaged volumetric fractions calculated with the correlation agreed well with the experimental data. Consequently, it was confirmed that the proposed model is a useful two-phase flow model.

View full abstract

In the present experiment, a single cavitation bubble is produced in a V-slot and the impulsive pressure is detected by four tiny pressure sensor elements. The size of a sensor element is 1 mm × 1 mm × 2 mm, and the four elements are set at 1.7 mm intervals along a line on the top surface of the pressure sensor. From the experimental results, the following is revealed. Impulsive pressure caused by microjet impingement in the V-slot is several MPa which is almost the same as that near the plane wall. When a bubble is produced near the bottom of the V-slot, the maximum impulsive pressure is detected at a point 3.4 mm from the point of microjet impingement, which is somewhat farther than the distance of 1.7 mm on the plane wall.

View full abstract

Because of further increase in waste heat due to the greater power capacity of equipment for use in space, the thermal management system using vapor-liquid two-phase flow has been attracting greater attention in recent years. This has led to a decrease in system weight and pump power as a result of improvements in the heat transfer performance. However, new information on the flow and heat transfer characteristics under microgravity conditions is required. In particular, the local distribution of void fractions in the cross section under the microgravity condition is expected to differ from that on earth. The flow behavior is investigated using an equal-density liquid-liquid mixture (silicone oil and water) as the working fluid in a horizontal pipe 25 mm in ID., which causes the effect of gravity on the two phases to be diminished due to buoyancy. In this paper, the experimental results related to flow pattern, void fraction, and pressure losses are reported, and compared with other investigators' results.

View full abstract

The buoyancy-driven motion of an encapsulated liquid drop in liquid-liquid-gas systems was analyzed experimentally and theoretically. To investigate the accelerated motion of the encapsulated drop in detail, an optical measuring device for the local upward velocity was designed. Furthermore, a simplified effective theory based on the experimental data was proposed for the rectilinear accelerated motion of the spherical encapsulated drop produced from the compound jet in the systems. The effects of drop radius and liquid physical properties on the accelerated motion were clarified numerically. The drag components, which were composed of the steady drag term, the added mass term and the Basset history integral term, were also analyzed. The numerical results of the upward velocity quantitatively agreed with the experimental results in the range of intermediate Reynolds numbers ; Re∼O(100).

View full abstract

Space-time correlations and power spectra of velocity fluctuations are measured in two-dimensional turbulent flows through curved and straight channels. Noticeable differences are seen between the experimental results of the two channels. While each effective power spectrum in the straight channel has only one maximum, each of those in the curved channel has two maxima on both sides of a wave number corresponding to the dissipation length. The spanwise correlation of radial velocity fluctuations in the curved channel takes a large negative maximum at an interval of about half the channel width. The streamwise integral length scales on the outer-wall side of the curved channel increase to about ten times those in the straight channel. These results indicate that intensive large-scale eddies which are unsteady but resemble Taylor-Gortler vortices exist in the curved channel.

View full abstract

This paper is the second in a series, which clarifies the effects of flow distortion by a sphere upon the statistical properties of a point-source diffusing plume evolved in grid-generated turbulence. In the preceding paper, only the results along the forward stagnation streamline were given. Here, various data of the entire concentration field around the sphere are investigated. It is found that the conditioned mean concentration is generally constant everywhere but in the recirculated wake region, but the conditioned concentration fluctuation rms values show a slight drop near the surface of the sphere in the nonrecirculating flow region. Moreover, except for the recirculated wake region, the conditioned probability density for near-zero concentration shows a slight drop, while that for the higher concentration increases slightly as the sphere is approached. Furthermore, from analysis of the spectra, the intermittency factor and the probability density distribution, it is concluded that only in the region next to the surface of the sphere have the molecular diffusion fully advanced and the concentration fluctuation dropped to nearly zero.

View full abstract

A numerical technique for computation of two-dimensional turbulent flows through an oscillating cascade is presented. To consider the interblade phase angle, a time phase shifted boundary condition is introduced at the periodic boundaries. A finite volume method is used to obtain the spatially discretized governing equations, while the Adams-Bashforth method of second-order accuracy is employed for the time integration. Furthermore, a two-equation model of turbulence is introduced to estimate the turbulence effect. To verify the effectiveness of the present method, computations are carried out for the flow through cascades of flat plate blades (inviscid flow analysis), lens-type blades and compressor blades. The present method gives unsteady periodic flow fields including the periodic aerodynamic force and moment acting on the blade clearly, and the negative damping force is obtained from the computational result of the flat plate cascade flow.

View full abstract

The flow characteristics of turbulent jets with and without flame were studied at the low Reynolds numbers of 2000 and 3000, with fuel-rich mixtures (equivalence ratios of 6, 24 and ∞ ) of propane gas and air. Axial and radial time-mean velocities and fluctuation velocities have been measured by LDV. Temperature and luminosity have been measured by a fine thermocouple and a photomultiplier, respectively. Potential core length was presented by normalized mass flux along the axis, and by radial mean velocities in the flow with and without flames. The potential core length in the flow with flame is about twice that in the flow without flame. The turbulent intensity, measured to study the effect of charged soot, in the highly luminous soot region was smaller than that in the nonluminous region, in spite of the same temperature and the same velocity gradient. The radial displacement of particles caused by the repulsive force between charged soot induces an increase in apparent gas viscosity in the flame gas. Order estimation of the viscosity by the movement of soot was carried out using electrical mobility and the equation of motion of particles in the shear region.

View full abstract

An experiment and theoretical study of a pulsating flow of magnetic fluid in a pipe in the case of applied fluctuating magnetic field are carried out. The pressure fluctuation in the pipe is measured and compared with the analytical results which are obtained by taking the particle aggregation into account. The experimental results show that the pulsating flow characteristics depend on not only frequency and fluctuating flow rate but also pipe radius. It is also clarified that the existence of a steady flow component reduces the rate of increase in fluctuating pressure difference due to the application of a pulsating magnetic field.

View full abstract

Study of unsymmetrical supersonic flow is important not only for basic research in gas dynamics but also for applications such as to jet propulsion and analyses of turbomachines. In the present study, a shock tube was used to generate supersonic flows in an unsymmetrical duct. The generation and oscillation of a pseudo-shock wave due to the interaction of a supersonic expanding flow with a boundary layer developed on the duct wall were studied experimentally using pressure transducers, and also by taking schlieren or schlieren interferometric photographs. The observed pseudo-shock wave oscillated in the flow direction. Its oscillation frequency was determined from schlieren photographs and compared with that obtained from pressure measurements. In addition, the Navier-Stokes equations were solved numerically to obtain steady-state solutions of the corresponding flow field.

View full abstract

Under low-speed operation the frictional characteristics of a piston with hydrostatic pads and a conventional piston were experimentally obtained and compared with the theoretical results previously reported. The metallic contacting force under medium- and high-speed operations is given theoretically. Finally, the effect of conically arranged cylinder bores on reduction of the radial loads acting on pistons is discussed.

View full abstract

The motion of capsules was numerically analyzed using the method of characteristics. It was found that the velocity of a capsule with large mass is lower than that with small mass, and that collision occurs when capsules of different masses are loaded into pipeline under a lower inlet pressure head. Time histories of pressure were also examined near the capsule loader and near the outlet of the pipeline. To confirm the numerical analysis, measurements in a horizontal pipeline were carried out using capsules of different masses. The numerical solution shows good agreement with the measurements.

View full abstract

A new analytical approach for optimizing shapes of the flow field is presented. The reshaping is accomplished by the growth-strain method which was first developed using the finite-element calculation of the deformation of shapes by generating bulk strain for solid problems. The generation law of the bulk strain is given as a function of a distributed parameter to be made uniform. For solid problems, the validity of the use of the shear strain energy density to maximize the strength based on the Mises criterion or the strain energy density to maximize the stiffness for the distributed parameter has been confirmed. In the present paper, we propose the use of the dissipation energy density for the distributed parameter to minimize the total energy dissipated due to the viscosity of the fluid. Numerical results for abruptly enlarged channel problems in steady state assuming low Reynolds number and incompressible viscous fluid shows the validity of the present approach.

View full abstract

Using the molecular dynamics method, the process of heat conduction in solid materials has been numerically studied for the configuration of lattice particles having the Lennard-Jones (12-6) potential. Under the condition of constant heat flux, the temperature distributions through the conduction layer are calculated with changing the amount of heat flux, layer temperature and layer thickness of the lattice particles, and the lattice configuration including defects. If averaged over sufficiently long time intervals, the molecular process of heat conduction exhibits the same features of Fourier's law as the macroscopic heat conduction. For less than four particle layers, the boundary conditions take part in the process of heat conduction, changing the relationships of the heat flux to the temperature gradient. Increasing the average temperature of the lattice particles deteriorates the heat transfer through the particle layer due to the increase in the temperature fluctuation of the particles. Lattice defects also greatly resist heat conduction, being associated with a large temperature gap across the defect.

View full abstract

This paper reports experimental results of natural convection heat transfer to liquid helium in a rotating field in the range of 1600 - 3700 rpm. Test surfaces are vertical plates (10 mm width, 5 mm height) with rectangular channels in a radial direction and upward- and downward-facing horizontal plates (8 mm diameter) at radial distances of 0.3 m. The results show that the heat transfer for vertical plates is approximately the same as that for horizontal ones facing upward, independent of the channel gap size in the range of 0.4 - 5 mm and direction of the Coriolis force. The heat transfer can be predicted from the conventional correlation Nu =0.13Ra^1/3 in the Ra range of 10¹¹ - 10¹³, if the film temperature for evaluating properties is substituted for the pseudocritical temperature only when it is above the pseudocritical temperature. The heat transfer for the horizontal plate facing downward is 1/6 - 1/7 of that for the one facing upward, and the Ra exponent of its correlation is 0.2 in the Ra range of l0¹² - 10¹⁴.

View full abstract

Heat transfer in natural-convection film boiling with a long vapor film was investigated experimentally. Experiments of film boiling on horizontal cylinders of large diameter and long vertical cylinders were conducted to investigate the effect of liquid subcooling on the film-boiling heat transfer. The test liquid was R-113 at atmospheric pressure, and the following results were obtained in the experiments. For cylinders of both types, local heat-transfer coefficients of saturated and subcooled film boiling were much higher than the respective predictions from the two-phase boundary layer theory. In the case of vertical cylinders, the local heat-transfer coefficient did not depend on the distance from the leading edge even if the liquid was subcooled. In the case of horizontal cylinders of large diameter, the local value did not strongly depend on the distance from the stagnation point for saturated film boiling, but it depended on the distance and reached a maximum at a distance from the stagnation point with increasing liquid subcooling. As for the averaged heat-transfer coefficient of horizontal cylinders of large diameter, it increased with increasing liquid subcooling but did not depend on the diameter even under subcooled conditions.

View full abstract

Film-boiling heat transfer of saturated methanol with decomposition reaction is investigated experimentally using a cylindrical heater with a catalytic surface. Platinum black powder which is bound to the cylinder shows sufficient activity to catalyze the reaction. The heat transfer coefficient markedly increases with slight increase of the wall superheat, becoming about 50% higher than Bromley's prediction for the case of no reaction. Such high heat transfer is considered to be due to the endothermic action and the evolution of hydrogen, the thermal conductivity of which is about five times that of methanol vapor. However, once the wall superheat exceeds about 510 K, the heat transfer coefficient decreases and approaches that for the case of no reaction. This may be due to the sintering of the catalyst in the present study. The heat transfer coefficient predicted by taking account of the absorbed heat and change of the properties of the vapor phase due to the evolution of decomposed gases is in good agreement with the high coefficient measured experimentally.

View full abstract

This paper reports a correlation of dew- and bubble-point curves for seven binary refrigerant mixtures of engineering importance. The present correlations have been established based upon the reliable measurements of dew and bubble points by the present authors and others using a constant volume method. We have made the present correlations by calculating the differences between the experimental and the calculated values using Raoult's law. After correlating these differences, ΔPd orΔPb with temperatures, we represented these temperature terms in terms of the composition χ₁ of the first components, which have the lower critical temperature of the binary components. It was confirmed that the present correlations for seven binary refrigerant mixtures effectively modeled the dew- and bubble-point curves for any mixture with an arbitrary composition, and reproduced the dew- and bubble-point pressures even in the neighborhood of the critical points where the Soave-Redlich-Kwong equation could not reproduce the measured vapor-liquid equilibria.

View full abstract

Laser Rayleigh scattering was applied for remote, nonintrusive point probing of the vapor concentration and its fluctuation near the spark plug of a motored spark ignition engine, which was caused by the intermittent injection of Freon-12 vapor into an intake port. The theoretical analysis showed that the instantaneous fuel vapor concentration can be split into three components : the ensemble-averaged mean concentration, the cyclic variation of temporal mean concentration and the temporal concentration fluctuation in a specific cycle. This paper concentrates on the measurements of the ensemble-averaged concentration fluctuation and the cyclic variation of temporal concentration fluctuation. The results show that the cyclic variation of temporal concentration fluctuation is too large to be neglected, and both the ensemble-averaged concentration fluctuation and the cyclic variation of temporal concentration fluctuation increase and reach a peak after which they decrease during the intake and compression strokes. They are strongly affected by air fuel-ratio, engine speed, injection duration and injection timing.

View full abstract

Laser Rayleigh scattering was applied for remote, nonintrusive point probing of the vapor concentration and its fluctuation near the spark plug of a motored spark ignition engine, which was caused by the intermittent injection of Freon-12 vapor into an intake port. The theoretical analysis showed that the instantaneous fuel vapor concentration can be split into three components : the ensemble-averaged mean concentration, the cyclic variation of temporal mean concentration and the temporal concentration fluctuation in a specific cycle. This paper concentrates on the measurements of the ensemble-averaged mean concentration and the cyclic variation of temporal mean concentration. The results show that both the ensemble-averaged mean concentration and the cyclic variation of temporal mean concentration increase and reach a peak, after which they decrease during the intake and compression strokes. Moreover, they are strongly affected by air-fuel ratio, engine speed, injection duration and injection timing.

View full abstract

The characteristics of engine∼load systems are considerably nonlinear. Therefore, when feedback control is required in such a case as a diesel engine governor or cruise control for a gasoline engine, a countermeasure such as gain scheduling must be incorporated to compensate this nonlinearity. On the other hand, it is said that fuzzy control is very robust against nonlinearity. Thus, fuzzy control was applied to the governor of a diesel engine, and the level of robustness of fuzzy control was investigated in comparison with PI and LQI control methods. As a result, it was found that fuzzy control showed the highest level of robustness and a much faster control response compared with the two other methods of control.

View full abstract

An experimental investigation was conducted on the structure of turbulent diffusion flames of hydrogen stabilized on a bluff body. A coaxial jet diffusion flame was formed using a cylindrical nozzle with an extremely thick rim, which works as a bluff body. In the present study, special attention was paid to the effect of combustion on the aerodynamic processes. The following results were obtained. (1) The combustion markedly enhances the penetration of the central fuel jet. (2) The laminarization phenomenon, reported in simple jet diffusion flames, was observed more conspicuously in this combustion field as well, and it was assumed to exert an important influence on the flame structure. (3) An accelerating flow field often exists in complicated flow fields with a reverse flow, where turbulence generation terms containing Reynolds normal stress become negative. This phenomenon sometimes exerts considerable effects on the turbulence behavior.

View full abstract

We observed the interaction between turbulence and combustion near the lean flammability limit of a premixed flame supported by a hot burnt-gas jet, and elucidated the detailed structure of flames varying with the Damkohler number. A typical wrinkled laminar flame appeared when the equivalence fuel-to-air ratio of the mixture was 0.6, whereas, for an equivalence ratio below the lean flammability limit (e.g., Φ_m=0.4), a structure which resembled a distributed reaction zone was observed. The latter showed features similar to those of the ordinary distributed reaction zone flame, which was observed in a strongly turbulent shear layer, in many aspects of the observed statistics of fluctuating properties, such as temperature, ion current, and radical chemiluminescence.

View full abstract

The flame ignition process of premixed methane-air mixtures on a high-temperature plate was numerically studied by considering 18 species with 61 elementary reactions in a laminar flow. The ignition is mainly controlled by the diffusion phenomena of heat and mass as well as the chemical reactions. The ignition is delayed at the same surface temperature as the equivalence ratio increases from 0.6 to 1.2, being mainly controlled by chemical reactions and only slightly influenced by the main flow velocity from 1 cm/s to 100 cm/s. Decreasing the surface temperature causes the ignition delay. The ignition delay time depends on whether the plate surface is chemically inert or deactivated for active species. Among active species such as H, O, OH and HO₂, H has the greatest effect on the ignition delay time. The ignition delay time increases exponentially as the area of hot surface is decreased.

View full abstract

The temperature stratification of the heated air flow in a fire tunnel is investigated using the two- and three-dimensional experiments along with numerical models described in previous papers. It is found that the heated air flow can appear to be either a stratified or a mixed flow pattern in accordance with different ventilation velocities, heat release rates, and downstream positions. To deal with tunnel fire circumstances, the mixed convection is evaluated using a nondimensional variable Gγ^*/Re^5/2_H. A quantitative relationship between Gγ^*/Re^5/2_H and a parameter which quantifies the varying degrees of the stratification at each tunnel cross section was established. By comparing the results from two- and three-dimensional analyses, the effects of secondary flow on the temperature stratification are discussed. The conclusions obtained are directly related to the safety considerations in tunnel fires.

View full abstract

In this study, we examine experimentally the dehydrogenation process of particles. Acetylene and propane soot taken from a diffusion burner and soot from liquid fuels (benzene, n-butyl alcohol and petroleum benzin) sampled from a wick burner are examined. To collect soot particles of different sizes, the sampling station is shifted along the centerline of the burner. The size distribution of soot particles is measured using an electron microscope. Carbon/hydrogen ratios are determined with a CHN analyzer. The measured carbon/hydrogen ratios plotted against the mean diameter D of the sampled soot fall on a straight line passing through the origin. This indicates that the number of carbon atoms grows in proportion to D³, whereas that of hydrogen is proportional to D². It is thereby concluded that hydrogen is located only on the surface of soot particles.

View full abstract