The lattice gas automata method was used to simulate fluid motion. In 7-velocity models, it is known as a ILG model was proposed as one of the models which can be manage two fluid components. Although the ILG model of 7-volocity model meets Laplace's formula, the pressure is independent of the density in 9-volocity model. In the present study, the ILG model was applied to 9-velocity model and collision rule was improved to handle the interface energy transport. As a result, it was clarified the improved 9-velocity ILG model meets not only Lapalce's formula, but also can deal with difference of density and heat transfer at the interface

View full abstract

This paper is concerned with a numerical method and its application to the prediction for turbulence structure which is induced in three-phase flows containing bubbles and particles. The Eulerian-Lagrangian model for three-phase flows has been constructed since the mathematical validity of the model was confirmed for bubbly two-phase flow in previous papers. Two-and three-dimensional predictions were carried out and provided detailed phenomenological information on three-phase interaction processes. When the bubble and the particle volume fractions were less than 5%, generation of kinetic energy was reduced more in three-phase flow than in two-phase flow, because of decrease in body force spectrum. Especially, the turbulence with long wave length was not strongly grown because of cancelation of meso-scale body force due to local dispersion of the bubbles and the particles. Resultant slope of energy spectrum in high wavenumber region became much calmer than the similar phenomena in bubbly two-phase flow cases.

View full abstract

The natural convection problems in a square cavity have been analyzed as a bechmark problem for two dimensional thermal fluid flow analyses. In this paper, the fouth-order weighted average finite difference method(FWA(C)is used for the problems of Rayleigh numbers from 10⁹ to 10¹¹. The FWA(C) algorithm is a combination of(a centered-difference)×W and(an upstream-difference)×(1-W). Here the W is a weighting parameter. Then, the FWA(C)is corrected by an artificial viscosity using the error analysis approach to improve its accuracy and stability. Non-uniform meshes are used to obtain a high accurate solution near the boundary. In these analyses, converged solutions are obtained for Rayleigh numbers from 10⁹ to 10¹¹. Thus, the effectiveness of FWA(C)was confirmed by comparing our results with other research works.

View full abstract

In fluid engineering, it is important to understand velocity and temperature profiles of target fields. Usually CFD or OFD techniques are used for understanding flow fields except for simple flow field which can be solved analytically, and they have basically been used independently even if they are sometimes compared mutually. But each technique contains merits and demerits. For example, CFD has inevitably errors accompanied by numerical nodeling and discretization, and on the other hand OFD includes lack of data and measurement error. In this research, we proposed a more rational technique that combines CFD and OFD techniques and overcomes these demerits by optimization technology using cost function. This cost function contains square of difference between observed and solution values, residual of continuity equation and N-S equation. We introduced some coefficients to determine the relative importance of these three terms, and conducted some sensitivity analyses. Then we obtained good results that agree with our intention and could verify the varidity of our cost function method.

View full abstract

To improve the measurement accuracy of particle image velocimetry(PIV), three or more images should be analyzed for the same particles. But the analyzing images were increased, calculation times increase rapidly. In this paper, a simple algorithm was devoloped to obtain flow velocity by using a personal computer. A lot of sequential image data are stored in large memory using a high-speed video camera. Those images are piled, and three-dimensional brightness distribution(x-y-t)data is prepared. These data are transformed to two dimensional image data by projection them to x-t or y-t plane. Then velocity components are derived from the derivatives of these two projected images, dx/dt or dy/dt. A standard image provided by the Visualization Society of Japan is analyzed to verify the algorithm and a measured result of separated flow around a cylinder is shown.

View full abstract

The enhancement of fluctuation of jet is very important in order to improve the mixing or diffusion properties with the surroundings. There are many ideas about this, and the use of resonance jet will be also one of them. But the details of the flow characteristics of resonance jet has not been made clear. In this paper the flow characteristics, such as resonance frequency, mean and fluctuating velocity profiles, spread of the jet and others, of the resonance turbulent jet and their control by changing the volume, length, of resonance room are made clear experimentally. As a result, the fluctuation and the decay of center line velocity of resonance jet are very large compared with conventional jet and others are made clear.

View full abstract

Effects of free and solid surface on round submerged water jets have been investigated experimentally. The jet is discharged parallel to a free and solid surface with an offset from each surface. Flow is visualized by using tracer particle method and the photograph obtained shows the present jets are classified into four types according to their behavior. Mean velocity of two kinds of flow, namely, the flow attaching to the free or solid surface is measured by a hot film anemometer. Characteristics of jets, such as the velocity profile, the decay and the location of maximum velocity, the half-value width and so on, are discussed on the strength of data. These results are also compared with the characteristics of free jet and the effects of both boundaries on the submerged water jets are clarified.

View full abstract

A circular jet from a nozzle set up center of the upper side disk is impinging on the lower side disk placed on a support. For the flow rate above some critical value, the lower side disk can be levitated according to Bernoulli's Principle. The dominating factors on the critical flow rate are the diameter and mass of the disk, d and m, and the space, s, the density and kinematic viscosity of fluid, p and ν, and the acceleration of gravity g. From the vector dimensional analysis, this phenomena is described by four dimensionless quantities: Q_c=q_c/dν, M=m/pd³, S=s³g/ν² and D=d³g/ν². The critical flow rate is given by Q_c=CM^0.35S^0.07where C=100 for d=40 to 75mm and C=90 for d=20 to 30mm, and the both errors are within ±4%. And the falling critical flow rate is given by Q_cf=33 M^0.208 D^0.09and the error is within ±4%.

View full abstract

This paper presents an effect of riblet shape on turbulent drag reduction. The experiment was carried out in an N.P.L.blow-down-type wind tunnel with a working section of 500mm×500mm×2000mm. In order to investigate the details of drag reduction for riblet shape, rib models of scaling up riblets were used in this experiment. The parameters were 5 kinds for the rib model and 7 kinds for the riblet model. The time-mean velocity and skin friction were measured by the Pitot and static pressure tubes and Preston tube respectively. The Reynolds stress was obtained using the data processing system connected to a hot wire anemometer. Consequently it was found that(i)the Reynolds stress and skin friction become smallest near the surface for the rib with the section of an equilateral triangle, and(ii)the drag reduction attains maximum for the riblet of V groove with nearly equilateral triangular section.

View full abstract

The linear stability analysis has been performed at Re_τ=180 for a turbulent-channel-type base flow with a periodic undulation in the spanwise direction in order to elucidate the generation mechanism of streamwise vorticity through the instability of streaks in near-wall turbulence. It is found that there appear three different instability modes depending on the spanwise wavenumber of streaks. In the case of the streak with around 100 wall-unit spanwise wavelength its critical velocity amplitude lies at 〓U⁺_c〓3, above which streaky flow is unstable to an infinitesimal sinuous disturbance, i.e.a bending mode along the streamwise direction. The instability is identified to originate from inflection points, i.e.wake-like instability, in the spanwise variation of the streaky flow. In this case, unstable eignmodes take a form that is inclined towards the streamwise direction from the wall-normal direction, and they directly induce the streamwise vorticity on low-and high-speed streaks. In addition, the streamwise vorticity is secondarily produced pricipally through tilting of the wall-normal disturbance vorticity by the base flos whear across the wall-normal direction.

View full abstract

In the previous papers, the swirling flow field in a rotating hollow turbine shaft was solved using computational fluid dynamics. It was observed that a large-scale spiral vortex existed at a place where the swirling flow turned radially outward. In this report, the pressure fluctuations in the swirling flow field are measured. The main part of the internal cooling air system of a gas turbine is used as the experimental apparatus. A specially devised liner in inserted inside the hollow turbine shaft and ten pressure sensors are embedded axially and circumferentially in the liner to measure the unsteady wall pressures. The pressure fluctuations which have the same characteristics as the rotating spiral vortex predicted in the numerical results are captured. The amplitude is great at the sensors near the place where the vortex was predicted in the nuberical results and the precession frequency of the rotating spiral vortex is in close agreement with the calculated frequency.

View full abstract

Stability of the surface of a magnetic fluid under a horizontal non-uniform alternating magnetic field is investigated experimentally and analytically. Several modes of interfacial waves of the same frequency as that of the applied magnetic field are observed, and the critical strength of the magnetic field at each frequency were obtained. The linear stability theory shows that the system is destabilized with the increase in the magnetic susceptibility and the gradient of the magnetic field, which is in accord with the experimental result. The iron nitride magnetic fluid, which is newly developed and has very high susceptibility and magnetization, was tested under the same condition as W-35 and HC-50. The small disturbance occurred even at very weak magnetic field, and the steady wave was formed in the uniform magnetic field.

View full abstract

The surface waves of a magnetic fluid under a horizontal non-uniform alternating magnetic field is investigated experimentally and analytically. While the frequency of the wave on a magnetic fluid is usually the same as that of the applied magnetic field, in the case of the iron nitride magnetic fluid, which is newly developed and has very high susceptibility and magnetization, the frequency of the wave is twice that of the applied magnetic field. This is confirmed by the linear stability theory. The shape of the surface, the motion inside of the fluid and the disturbed magnetic field are presented.

View full abstract

Turbulence distortion by solid particles in a fully-developed channel flow was investigated by Lagrangian measurement technique. Digital particle image velocimetry and a high-speed CCD camera mounted on a shuttle moving with particle mean streamwise velocity were used to simultaneously detect particle and fluid information amongst particles. Two classes of particles at particle volumetric fraction up to 3.32×10^-4 were employed to provide characteristic values of length ratio and distance between particles, i.e., inter-particle spacing. Turbulence augmentation by particles greater than the Kolmogorov micro length scale was induced by the region of the high-strain rate and high vorticity on both sides of particles. The Rapid Distortion Theory applied to dispersed two-phase turbulent flows indicated that the directional non-isotropic structure was observed with an increase in particle volumetric fraction. The enstrophy production was enhanced with decreasing values of inter-particle spacing, yieding an increase in the enstrophy amongst particles.

View full abstract

The present study is concerned with experiments and analyses of the movement of a liquid drop in a horizontal circular pipe exposed to air flow unjected from a vertical circular pipe. Each inner diameter of the pipes is 30mm and both ends of the horizontal pipe are open to the atmosphere. In the first experiment, a water drop is softly put on the inner wall of a horizontal acrylic resin pipe and its shape is examined. The approximate drop shape can be calculated from the law of minimum energy and the information obtained about the contact angle. In the second experiment, a water drop is exposed airflow on a curved wall. It is clarified that this phenomenon is closely related to the gas flow in the horizontal pipe. And the relation between air flow and the stop position and the split-phenomena of liquid drop are examined experimentally and analytically.

View full abstract

In the present paper, the effective diffusivity of gas exchange in the 1st and 2nd bronchial generation airways has been experimentally studied in oscilatory flow through the bifurcating tubes of verious diameters. And then we made an attempt to evaluate theoretically the effective diffusivity in the bifurcating tubes on the basis of the solution for a straight tube. The respiratory condition in this study corresponds to the High Frequency Ventilation.

View full abstract

The discharge coefficients of the sonic Venturi nozzle were measured for eleven gases on the Reynolds number range from 2×10³ to 4×10⁴. The results showed that the discharge coefficients strongly depend on gases. And it was also suggested that the discharge coefficients of most of the gases tested can be described by using two parameters theoretically determined on the assumption of isentropic flow of ideal gas, if they are on the conditions which are not so far from the ideal gas state. The differences between the theoretical and the experimental discharge coefficients were within 0.5 percent, except for CO₂, SF₆ and C₃H₈.

View full abstract

Structure and velocity measurements using magnetic resonance imaging(MRI)have been performed experimentally to obtain a correlation between pore structure and interstitial flow through the packed bed of 5mm diameter in the tube of 36mm ID. To measure axial velocity maps of water flow through the packed bed, the phase method of using the phase difference of water spin magnetization between flowing and atagnant fluids by applying magnetic fields with bipolar gradients was employed. The spatial resolution of the obtained map in 0.2mm×0.2mm×0.5mm. It was made clear from the obtained axial velocity maps that channel flows with higher axial velocity were induced not only near the wall but also in the internal region of the packed bed. Furthermore, pore structure of the packed bed was characterized from multi-slice images by partitioning of void space and combining of each pore section along the axial direction to analyze the structure-flow correlation. It was found from image analysis that axial channels with long and straight void space existed in the pore structure, and that most of the channel flows with higher axial velocity were induced in the axial channels. The flow rate through an axial channel depends on the square of the averaged cross section of the axial channel.

View full abstract

The aerodynamic and the noise characteristics of a centrifugal fan were obtained experimentally. Typical velocity components-for example absolute velocity and relative velocity-were caluculated from the characteristics of aerodynamics and the theory of fans. Changes in the typical velocity components with changes in the flow coefficient were converted to the change in the noise level by using 6th dependency of the power to the velocity and compared with the characteristics of noise. The results showed that the noise and the aerodynamic characteristics were closely correlated and showed that the principal velocity determining the noise level was relative velocity. A model for predicting the noise level of a centrifugal fan was described and its predictions were in good agreement with the measured levels.

View full abstract

In order to control and suppress a performance-curve instability characterized by a local dent in head-capacity curve or an uprising head-capacity curve in a mixed flow pump, an innovative and very simple device of using shallow grooves mounted on a casing wall is proposed. The optimum dimension of the groove and its location for the instability supression are determined experimentally. The results show that the shallow grooves located with proper dimension can sufficiently suppress the instability for all discharge ranges. The remarkable effect of shallow grooves is to decrease both the swirl strength and the size of the reverse flow region at an impeller inlet by use of angular momentum absorption through the mixing between the main flow and the groove flow, resulting in the recovery of impeller theoretical head.

View full abstract

The relative and turbulent flows were measured at the mid-passage of an unshrouded centrifugal inpeller. The measurements using a hot-wire anemometer were made for a flow rate corresponding to nearly zero incidence angle and two other flows with reduced and increased flow rates. A low relative velocity, a high dissipation and a high Reynolds stress appear around the region of the middle blade-to-blade to the casing side due to both influences of a blade tip leakage and a secondary flow. As the flow rate is reduced, this region enlarges by nearing the hub and also tends to move toward the suction. In the increased flow rate, the Reynolds stress is remarkably high near the casing of the suction side since the tip leakage noticeably influences there. In the reduced flow rate, the stress becomes high also on the suction because of the secondary flow strengthened near the casing. For three flow rates, however, the dissipation is remarkably high near the casing of the suction side; the influence of the flow rate on the position with the highest dissipation seems to be little.

View full abstract

Performance of a Wells turbine was improved by equipping 2D guide vanes before and behind a rotor. For further improvement, 3D guide vanes have been proposed in this paper. The performance of the Wells turbine with 2D and 3D guide vanes have been investigated experimentally by model testing under steady flow conditions. Then, the running and starting characteristics in irregular ocean waves have been obtained by a computer simulation. As a result, it is found that both of the running and starting characteristics of the Wells turbine with 3D guide vanes are superior to those of the turbine with 2D guide vanes.

View full abstract

Wavy condensate films flowing on a vertical wall have been simulated with a finite difference method in which the algorithm is based on the HSMAC method and free boundary conditions are treated with newly proposed method. A random perturbation given at the leading edge quickly diminishes once, and propagates downstream as small-amplitude long waves. Then the amplitude of the wave increases rapidly at a certain position. The wave shape changes from a sinusoidal wave to a solitary wave which is composed of a big wave and capilary waves. A large circulation flow occurs in the big wave and temperature contour lines are distorted by the circulation flow. The heat transfer is enhanced by space-time variation of the film thickness and convection effects. The higher enhancement rate is obtained for higher Prandtl number. In the present simulation condition, the wave inception occurs at lower film Reynolds number region for higher Prandtl number.

View full abstract

Microwave heating of a dielectric in a cavity was analyzed numerically with the FDTD method with the aim of devising new methods for reducing uneven heating typical of microwaving. The dielectric was assumed to be water and the frequency of a microwave was taken to be 1 GHz. It was found that the electric field is highly dependent on the position of a dielectric in a cavity. The temperature distribution reflects the profile of the RMS value of the electric field in a dielectric, although there appears to be no effect of the short wavelength typical of the electric field. The heating rate depends on the position of a dielectric. In the case of higher effective loss factor a microwave decays immediately after entering a dielectric, and the temperature of the interior remains low. These results indicate that the uneven heating is due to at least two causes: standing wave and rapid decay of a microwave.

View full abstract

Melting heat transfer of ice in porous media around a single horizontal cylinder and two horizontal cylinders vertically spaced in a rectangular enclosure has been experimentally and numerically investigated. In the experiments, an ice saturated porous medium was set at-1.0°C, and the temperature of the cylinder surfaces was changed above the fusion temperature, 0°C, and then melting was initiated. Melting of pure ice without porous media was also studied in order to compare to the case with the porous media. As a result, it was seen that natural convection largely affected melting of pure ice without the porous media even for the small cylinder-temperatures, 4.0°C and 8.0°C, but with the small porous particles(d=1.0mm)the effect did not largely appear even for the large cylinder-temperature, 20.0°C. On the other hand, with the large porous particles(d=5.0mm)the effect is clearly observed for 20.0°C. The melt area ratios were not largely changed with a change in the number of cylinders, although the melt shapes were largely changed after melted regions around the two cylinders were merged. Numerically obtained temperature distributions and melt area ratios were in good agreement with the experimental ones. In addition, timewise variations of velocity distribution in melt were clarified by the calculations.

View full abstract

The generation mechanism of the vertical vortices associated with the baroclinic instability and the effects of nonlinear term the vortices are investigated by using both direct numerical simulation(DNS)and rapid distortion theory(RDT). Two kinds of the anisotropic flow fields are used as initial conditions. As a result, the initial anisotropy of Reynolds stresses is found to affect asymmetry of the vertical vortices. In the cases where the initial vertical velocity is set to be zero, the p.d.f.of the vertical vorticity tends to incline toward the anticyclonic side. When the vertical component of initial velocity is larger than the horizontal one, the cyclonic vortices are more enhanced. By comparing DNS and RDT, it is found that in both cases of the intial conditions the enhanced vortices of DNS are stretched in the vertical direction, which is not observed in the RDT results. This should be because the nonlinear vortex-stretching term intensifies and elongates vertical vortices in the vertical direction. The anticyclones are markedly augmented in low Prandtl number fluids, while the cyclones become dominant in the high Prandtl number case. In particular, the flow field becomes almost two-dimensionalized and Taylor columns are formed in the vertical direction in the low Prandtl number case. However, neither two-dimensionalization nor Taylor column is observed in the RDT analysis which neglects the nonlinear terms.

View full abstract

Experiments have been performed for a turbulent channel flow obstructed with a flat body. The local heat transfer coefficient and the wall static pressure were measured on the 2 kinds of flat bodies of which the trailing edge shape differed. Length of the body, thickness of the body and distance between the wall and the body were changed in several steps. The total performance between heat transfer and pressure drop was estimated under the conditions of an equal pumping power. The total performance hardly changed, even if the trailing edge shape and length of the bodies were different. The averaged heat transfer coefficient increased with increasing thickness of the bodies, however, the friction factor increased over it, and the performance became worse. When the comparatively thin body was installed near the heating surface, the good performance was obtained.

View full abstract

An analytical method has been developed for the inverse heat conduction problem, when the temperatures are known at two positions in a finite body. On the basis of these known temperatures, a closed form solution is determined for the transient temperatures beyond the two positions by using Laplace transform technique. The method first approximates the temperature data with a half polynomial power series of time. The expressions for the surface temperature and the surface heat flux are explicitly obtained in the form of the power series of time. Numerical results for some representative problems show that the temperature and heat flux can be predicted well by this method.

View full abstract

An experiment has been carried out to elucidate the critical heat flux(CHF)of an open two-phase thermosyphon with a bottom heated chamber in which heat is absorbed by evaporation of liquid. Another objective is to enhance the CHF using a concentric-tube by which counter-current flow of vapor and liquid in the throat of the chamber can be controlled well. The CHF data are measured for the saturated liquid of R 113 at a different pressure and different configuration of concentric tubes. The CHF data without the inner tube are in good agreement with the existing correlation and analytical result. The CHF increases by as much as several times of the CHF without the inner tube with an increase in the inner tube diameter up to a certain diameter of the inner tube and then decreases continuously as the inner tube diameter approaches the outer tube diameter. The optimum diameter of inner tube exists at which the CHF is maximum.

View full abstract

All the correlations and analytical models of critical heat flux in saturated pool boiling predict that the critical heat flux under microgravity conditions becomes much smaller than the terrestrial value. In the present paper, aiming at the enhancement of the critical heat flux under microgravity conditions, a novel technique to shorten the bubble departure period τ_d in saturated pool boiling under microgravity conditions was proposed. The technique proposed is based on the self round-off motion of bubbles induced by surface tension. Such self round-off motion is expected to occur by setting ribs of an asdequate height on the boiling surface. In the present experiment, to realize and simulate such a boiling system, saturated boiling experiments were conducted using a ribbon heater sandwiched by two glass plates. In the experiments, the glass plate height H was changed from 0.5mm to 20mm. Experimental data under microgravity conditions indicate the following results. In the cases of H=0.5-4.5mm, the bubble departure period of ethanol under microgravity conditions is as short as the terrestrial value and then the critical heat flux is as high as the terrestrial value. The critical heat flux is caused when the value of τ_dq reaches a maximum. The equivalent liquid-film thickness calculated from the experimental data of τ_dq is much smaller than the macrolayer thickness calculated from the model proposed by Haramura and Katto.

View full abstract

Most of the research on district heating-cogeneration system deals with fixed servic area and energy demand. This paper presents a computational method for detecting location and service area of a district heating network with cogeneration, wich offers the maximum cost reduction compared with the existing energy system cost. The existing energy system consists of boilers and electric power from power grid. In the method the area in interest(city or province)is divided into small geographical zones. Its heat and electric demand and the necessary geographical coordinates characterize each zone. The basic parameter in the computational method is the Cost Reduction Factor(CRF), which is defined as the cost reduction rate that can be achieved if the existing energy system is replaced by a district heating-cogeneration network. The aim of the model is to connect zones in group, named cluster, with positive CRF. Effects of pipeline cost, fuel and electricity costs and hourly energy demand patterns on CRF are all included. The program is applied to an example area, and changes in CRF and CO₂ emissions are analyzed for different size of clusters(network area).

View full abstract

A solid polymer water electrolysis system is a technology to make hydrogen from water by electric power. It uses solid polymer electrolyte nembrane, which has proton conductivity. This system has characteristic of low operation temperature, can make high purity hydrogen and oxygen, and needs only water and electric power. Electrolysis is electrochemical reaction with heat generation. So thermal design of large size module is very important to make high performance system, because its performance varies with temperature and it has upper limit temporature about 120°C. A solid polymer water electrolysis module is composed many stacked cells and separators. A performance simulator of solid water electrolysis module was made, considering three-dimensional distribution of tempeorature and current density. The influences of module size and operate conditions upon module performance were studied.

View full abstract

An experimental and analytical study has been performed to investigate the production behavior of slush ice by use of an oscillatory rotating cooled tube which was installed vertically in a test vessel filed with ethylene glycol solution. Experiments were carried out under a variety of conditions, such as angular acceleration of the cooled tube, rotation angle, cooled wall temperature, and the initial concentration of solution to determine the production performance of slush ice. It was found that a thin frozen layer which was formed on the cooled tube was continuously separated from its surface depending on the operation conditions, and as a result, a large amount of fine ice crystals were produced in the test vessel. The operation range in which the continuous production of slush ice was identified experimentally. Furthermore, an analysis was made to determine the characteristics of ice packing factor(I.P.F.)by use of an one-dimensional heat transfer model. The range of initial concentration of solution was determined where the tendency of the analytical results and experimental ones agreed well.

View full abstract

Experimental investigation was conducted on two droplet-array combustion of methanol and methanol/dodecanol mixture fuels in microgravity. For methanol, effects of ambient pressure and droplet spacing were examined. Results show that the droplet lifetime decreases with increasing spacing at relatively low pressure and the droplet lifetime becomes independent of spacing at higher-subcritical and supercritical pressures. For methanol/dodecanol mixture, effects of pressure, fuel composition were investigated in terms of occurence of disruption. Disruption of droplet during combustion was demonstrated both for single droplet and droplet pairs.

View full abstract

Using optical measurements and simplified thermodynamic models, thermal and fluid dynamic properties of the flame jets near the ignition limit are analyzed in a methane-fueled and constant volume prechamber combustor. Two types of cooling effects act on the flame jet; one is named "the convection cooling" by the nozzle wall during passage through the nozzle, the other is named "the mixing cooling" due to the entrainment of cold mixture after issuing into the main chamber. Model analyses show that, near the ignition limit, the convection cooling causes the flame jet a moderate temperature drop of about 200 K, and that the mixing cooling leads to a drastic drop in the flame jet temporature of about 800 K. Therefore, it is confirmed that the ignition process by the flame jet is severely affected by the thermal effect. Reasurements of CH-emission intensity from the flame jet at the nozzle exit also indicate that there exists almost on flame kernels in the flame jet weth the nozzle diameter below the ignition limit. The increase in the nozzle diameter enhances the thermal and chemical activity of the flame jet and leads to the early ignition, which in turn shortens the effective time of the flame jet. The shorter the effective time of the flame jet becomes, the slower the burning process in the main chamber. It is considered that the combustion acceleration is due to the effect of the flame jet on the stirring and dispersion of heat and active species over the entire main chamber.

View full abstract

As a means for a stable premixed combustion, there is a so-called cyclone combustor, which consists of a cylindrical chamber and fuel nozzles installed tangentially on the side wall. In this combustor an extremely stable flame can be obtained in the swirl flow, formed along the inner wall of the combustor. The authors utilized this combustor as a flame holder, to burn a high velocity jet flowing axially in the central part, and named this new combustor a cyclone-jet combustor. In the present study, the excellent flame stability is shown for the cyclone-jet combustor and, comparing among premixed, non-premixed and partially premixed flame, the low NO_x combustion characteristics was experimentally examined for this combustor.

View full abstract

To examine flame structure in turbulent flow, we have established highly tubulent flames of a propane-air mixture in a cyclone-jet combustor, and have measured ion current by an electrostatic probe. Based on the measured flow field by a hot wire anemometer, we have obtained turbulence characteristics in the non-reacting flow, and have plotted them on the Peter's phase-diagram to classify the micro-structure of turbulent flames. From the ion current characteristics and tomographic images of flames, we have discussed the flame structure and turbulence. Results show that, as the exit velocity is relatively low, a wrinkled laminar flame is formed. As the exit velocity is further increased, the flame wrinkling is intensified to increase the fluctuation in wave of ion current. When the velocity exceeds 30m/s, the PDF of ion current is no longer that of a wrinkled laminar flame, which shows that, with turbulence, one continuous flame sheet cannot be sustained, and the destruction of laminar flame structure occurs, in which small reacting eddies in an intermediate stage of combustion are embedded. It appears that the flame structure changes from the wrinkled laminar flame to those with distribution reaction zone.

View full abstract

The in-situ, quantitative profiles of temperature and OH radical concentration in a post-flame region between two-dimensional laminar counter-flow twin flames were measured by wavelength modulation spectroscopy(WMS)using a 1.5μm external cavity diode laser. The second harmonic('2f')signal was generated from absorption by two different overtone rotational-vibrational transitions of OH; the P 11.5(υ′, υ")=(2, 0)(ν₀=6421.354cm_-1)and the P5.5(υ′, υ")=(3, 1)(ν₀=6434.613cm_-1)trasitions. The absorption occurred in the post flame region between propane-air premixed twin flames stabilized in a two-dimensional laminar counter-flow burner(Tsuji burner)with a 60mm line-of-sight path length. The temperature and OH concentration profiles at different equivalence ratios(φ=0.8 and 0.9)were determined by least-square fitting of theoretical 2f lineshapes to the experimental counterparts and by calculating the ratio of the line intensities('2-line method')of the two different OH transitions. The measured temperature and OH concentration profiles were qualitatively consistent with two-dimensional numerical predictions using an augmented reduced mechanism of propane-air premixed combustion which includes OH as principal species.

View full abstract

The reactive mixing layer disturbed by three-dimensional turbulence issued from the splitter plate has been investigated numerically by solving the three-dimensional governing equations of mass, momentum, species and energy, to make clear the interaction between lifted diffusion flames and three-dimensional turbulence. The employed chemical kinetics is one-step, irreversible methane/air reaction under the Arrhenius law. The numerical results have shown three-dimensional turbulence inducing the flame stability. The tip of the flame base protrudes locally to make juts. The low velocity regions occur just before the flame juts leading to the flame stability. It is found that the low velocity regions are generated by vortices attributed to the baroclinic torque, based on the deviation of the density gradient and pressure gradient vectors, which is responsible for the geometry of the juts.

View full abstract

In this study, a new concept is proposed in order to simulate the behavior of multicomponent fuel spray impinging on a hot surface. Here, distillation curve of the fuel is considered to describe the vaporization property of each component in gas oil, since the superheating degree of the surface, defined by the temperature difference between the wall surface and the fuel saturated temperature, affects the solid-liquid boiling state. At the first, some impinging sprays structured by single component fuel are calculated to confirm the difference of vaporizing characteristics and impinging behavior in each fuel. Secondary, the spray impingement submodel is made. In this model, the difference of droplet breakup, reflection and dispersion process with the difference of boiling state is considered based on several experimental results of impinging droplet on a hot wall. This new submodel is incorporated into KIVA-II code. And the results calculated by this model are compared with that by KIVA-II original code.

View full abstract

Significant improvements in exhaust gas emissions and engine performance in an ordinary DI diesel engine were realized with highly oxygenated fuels. The smoke sharply and linearly decreased with increases in oxygen content regardless of the kinds of oxygenates to reach zero at an oxygen content of 38wt-% even at a stoichiometric condition. Diethylene glycol dimethyl ether(DGM) was used as a base oxygenate, and dimethyl carbonate(DMC)or methanol were blended with DGM at a volume ratio of 80:20 for perfectly smoke-free fuels. The NO_x, THC, and CO were almost all removed with a three-way catalyst at higher BMEP where the stoichiometric diesel combustion was realized. At lower BMEP with higher excess air the EGR effectively reduced NO_x and the three-way catalyst removed almost all THC and CO. The engine output with the highly oxygenated fuels were significantly higher than those of the conventional diesel fuel due to the release from smoke limit.

View full abstract

The partial flow dilution method is one of the typical measurement methods for particulate matter emission from diesel engines. In this method, exhaust gas at a transient flow rate should be transferred to a dilution tunnel at a constant ratio of exhaust gas. The present partial flow dilution method is used under steady-state engine operating conditions in lieu of direct flow rate measurement of exhaust gas. A more practical control of exhaust emission is, however, required world widely; therefore development of an exhaust gas flowmeter is indispensable in the partial flow dilution method for transient engine operating conditions. An ultrasonic exhaust gas flowmeter has been developed and been demonstrated to be capable of measuring the exhaust gas flow rate with sufficient accuracy.

View full abstract

In order to clarify the triggering condition for vapor explosion, experiments were conducted by passing pressure pulse through the film boiling on high temperature particle surface. The film collapse behavior was observed by a high speed video camera of maximum speed 40, 500fps. At the same time, time transient of the particle surface temperature and the pressure near the particle were simultaneously measured. As the results, the vapor film started to diminish just after the pressure pulse passing through the particle, then it was observed that many small splashes covered whole of the vapor film at the minimum vapor film thickness. Heat flux released from the particle surface increased with decreasing the vapor film thickness befoer the splashing and the heat flux almost constant after the splashing. It is supposed that the vapor film collapse can be judged by the visual information at the splashing. It was experimentally indicated that the measured pressure at the splashing is almost constant about 0.4MPa independent on the initial particle surface temperature.

View full abstract

Thermal detonation model is proposed to describe vapor explosion. According to this model, vapor film on pre-mixed high temperature droplet surface should be collapsed for the trigger of the vapor explosion. It is pointed out that the vapor film collapse behavior is significantly affected by the subcooling of low temperature liquid. However, the effect of subcooling on micro-mechanism of vapor film collapse behavior is not experimentally well identified. The objective of the present research is to experimentally investigate the effect of subcooling on micro-mechanism of film boiling collapse behavior. As the results, it is experimentally clarified that the vapor film collapse behavior in low subcool condition is qualitatively different from the vapor film collapse behavior in high subcool condition. In high subcool condition, instability of the vapor film dominates the vapor film collapse on the particle surface. On the other hand, micro-mechanism at the interface netween vapor and liquid such as micro-jet is dominant in low subcool condition in case of vapor film collapse by pressure pulse.

View full abstract