We have developed a numerical scheme to reproduce the unsteady flows with cavitation by the finite-difference method. The evolution of cavitation is represented by the source/sink of vapor phase in the incompressible liquid flow. The pressure-velocity coupling is based on the fractional-step method for incompressible fluid flows, in which the compressibility is taken into account through the low Mach number assumption. We applied our method to the cavitating flows around a hydrofoil. In the 2-dimensional simulation, the lift force was compared with experimental data and the reasonable agreement was obtained as for the dependence on cavitation number. In the 3-dimensional simulation, a cycle from surface oscillation of attached cavitation bubble, shedding of vortex and cloud cavity, and the re-entrant jet was successfully reproduced.

View full abstract

Measurements of turbulence intensities downstream from a 90 degree elbow of circular cross section in the case of 1.1, 2 and 4 radius ratios were obtained using a laser-Doppler velocimeter at Reynolds numbers of 5×10⁴-1×10⁶ in order to clarify the Reynolds number effect on turbulence downstream from the elbow. The cross-sectionally averaged turbulence intensity decreased slightly with increasing Reynolds number. The effect of Reynolds number on it near the elbow was smaller than that for fully developed turbulent flow and decreased with decreasing radius ratio. The power spectrum of the turbulence intensity downstream near the elbow with any radius ratio had a distinct peak at the reduced frequency of about 0.5. No marked effect was observed of the Reynolds number on it.

View full abstract

A numerical method for predicting submerged vortex cavitation in a pump sump is presented. In the present method, a global flowfield around vortices is determined by the help of numerical simulation of fluid dynamics. In many practical cases, the viscous core radius of a submerged vortex is very small compared to numerical grid spaces. To compensate for the lack of vortex resolution in the global solution, a vortex model, which is the exact solution of Navier-Stocks equations in a stretching vortical flow, is applied to the local flowfields around analytically identified vortex positions. Using circulation and axial velocity gradient calculated from global flowfields and prescribed fluid viscosity, the minimum pressure in the vortex core is calculated. It is found that the predicted values agree well with the experimental ones in terms of a critical inlet velocity at cavitation inception.

View full abstract

A numerical simulation technique of creature's flow was proposed on the basis of molecular dynamics method. Creatures, such as humans, fishes, bird and so on, usually move in a cluster or crowd. Features of crowd's behavior were discussed from the point of fluid dynamical view. The motivation of its motion was classified into a several categories due to intelligent level. In the present report, three primitive categories were taken into the consideration. In order to express a suitable individual potential, directional and influence functions were introduced to with a dependency of creature's view angle. Macroscopic quantities of state such as temperature, density, and pressure of the crowd were defined for a local region in the creature flow. An interaction between two intellectual particles and crossing flows between two crowds were simulated and discussed in detail.

View full abstract

For a compressible fluid under gravitational force, the density stratification is not enough and the entropy stratification is essential. In this report, we present the numerical simulation of a line sink compressible straified flow in two-dimensional channel by the finite difference lattice Boltzmann method. The results show that selective withdrawal is establithed when the entropy increase upward that is predicted by theoretical consideration.

View full abstract

A parallel LES technique for viscous incompressible flow with heat transfer based on the finite-element method using tetrahedral elements has been developed. It uses an algorithm based on the SIMPLER method to solve the Navier-Stokes equations, in which a BTD term is introduced to ensure computational stability. In this technique, the Crank-Nicholson method is used as a time integration scheme to improve the accuracy to solve both the Navier-Stokes equations and the energy equation. Moreover, the Smagorinsky model is applied to approximate the Reynolds stress term and the zero-equation model is applied to solve the energy equation. The code based on this technique is parallelized by using mesh partitioning by finite elements in order to carry out the element-based calculations and by using row-based domain decomposition of the global matrix to solve the linear equations. The developed LES technique was verified by simulating natural convection problems in a cubic cavity at Ra=10³ to 10¹0. Numerical results at Ra=10³ to 10⁸ agree well with those computed by other researchers and successfully demonstrated the stable prediction of the unsteadiness and the heat transfer at Ra=10⁹ to 10¹0. It was also applied to a realistic heat transfer probelm in a rotating cavity. Computed heat transfer coefficients at Ra=1.09×10⁹ and 2.18×10⁹ were found to be within 15% of those measured experimentally. This acceptable error means that the new LES technique can be applited to solve heat tranfer problems in a rotating flow -field.

View full abstract

The lateral flow force acting on the spool was analyzed by numerical simulations of three-dimensional flow. The test spool valve was the practical hydraulic type having non-axisymmetrical flow paths with inlet and outlet ports near the valve orifices. In case of the inflow jet, lateral flow force was found to depend on the jet impinging pattern such as jet angle and jet impinging point in the azimuthal direction on the surface of spool. The sum of the lateral forces is very small because the forces acting on the surface at the different axial positions counteract each other. In case of the outflow jet, the force is dependent on the jet impinging pattern as well as the flow pattern near the outlet port edge. The sum of the forces acting on the sorface in the same direction increases with decreasing the spool stroke.

View full abstract

Chemical vapor deposition (CVD) process composes a complex system, where chemical reaction and heat and mass transfer interact with each other. And these macro-scale phenomena are deeply related to micro-scale mechanics. In this paper, we present the macro-scale simulation by the DSMC method. In those reactors, sometime the important species such as the reactive intermediates have extremely low density ratio. This causes the large statistical fluctuation in the DSMC method, where the number of particles and the calculation time are limited. We propose a new numerical method for this kind of problem and the whole process of silicon CVD is simulated by the new method. It is confirmed that the new method is very effective and make it possible to analyze the CVD process more precisely.

View full abstract

The objective of the present study is to obtain fundamental knowledge of the influence of the lift force on the bubble and its motion induced by the surrounding flow field. We explored the flow structure in the vicinity of the bubble and also deformation of the bubble shape respectively by PIV/LIF and a projecting technique. As our system, we chose a single air bubble with an equivalent bubble diameter (D_eq) 1∼8 mm, in a vertical shear flow. Velocity measurements were made using one digital high-speed CCD camera for Digital Paricle Image Velocimetry (DPIV) with fluorescent tracer particles. The recorded image data are analyzed by cross-correlation technique. A second CCD camera was used to detect the bubble's shape and translational motion via backlighting from a square array of infrared LEDs. We confirmed that the bubble deformation and surrounding flow induced lateral force on bubble. We proposed the lateral force parameter αr which consisted of the relative velocity of the bubble and the circulation around the flow structure in the vicinity of the bubble measured by PIV/LIF. Bubble motion can be estimated by αr with the phase which depends on the circulation around the bubble.

View full abstract

The cold flow experiments are made in a two-dimensional dump diffuser with a split flow entrance at flame dome head. Flow behavior over a flame dome head and its reattachment to the inclined wall are presented, when a symmetric flow is provided at the diffuser inlet. A Laser-Doppler Velocimeter is used for the velocity measurements. The effects of four types of inclined wall in a sudden expansion region on the diffuser performance are clarified. Throat area between inclined wall and flame dome head is the dominant parameter affecting the total pressure loss.

View full abstract

Numerical investigations have been made of the oscillatory entrance flow in a straight pipe connected perpendicularly to the wall of a large reservoir. The pipe opening forms a bell mouth, the roundish corner radius of which is the same as the pipe. Calculations are carried out under the conditions of the Womersley numbers of α=2∼20 and the Reynolds numbers of Re=100∼500. The velocity profiles, the streamlines and the distributions of the wall shear stress are illustrated. The influences of the Womersley number and of the configuration of the pipe opening on the entrance flow phenomena are made clear, as occasion calls, by referring to results for a pipe with a sharp edged opening. The flow near the inlet in the outflow as well as the inflow terms exhibits a complicated behavior depending on the Womersley number. The separation at the bell mouth in the outflow term comes to be suppressed with an increase of α. In place of that, for moderate and high α the vortex flow is formed just in front of the bell mouth at the phase that changes from the outflow to the inflow terms. In consequence of the developing of the axial velocity along the pipe axis, the one-directional or bi-directional steady steaming is induced in the entrance region. The inlet length determined from the amplitude of the axial velocity is appreciably shorter than that for a steady flow and it changes proportionately with Re/α², except for sufficiently high α.

View full abstract

This paper presents model test results for the hydrodynamic pressure caused by the streamwise and vertical vibrations of the vertical flat weir plate, to verify the theoretical analyses for hydrodynamic pressure, developed by Ishii and Anami, for its validity. The streamwise vibration pushes and draws the upstream water, thus inducing a excessive push-and-draw pressue, while the vertical vibration fluctuates the flow rate under the weir plate, thus inducing a flow-rate variation pressure. A vertical flat weir plate model standing over the horizontal channel floor was forced to vibrate at a small amplitude, in the streamwise and vertical directions, respectively, and the vibration-induced hydrodynamic pressure was measured along the weir plate. The measured data were compared with the theoretical ones, to show their close agreement.

View full abstract

This paper describes the fluid force acting on the two-dimensional oscillating square cylinder in the locking-in region. The experiment was carried out in an N. P. L. blow-down type wind-tunnel with a working section of 400 mm×400 mm×2 000 mm, and with the Reynolds number of 1.4×10⁴. The cylinder was then forced to oscillate sinusoidally in the lift direction. The power spectrum in the wake behind a square cylinder was measured to show the locking-in region in the present experiment. The time-mean pressure distribution and fluctuating pressure distribution on the surface of the square cylinder were measured for the displacement in the vibration. Consequently it was found that the mean drag and fluctuating lift increase and become maximum in the locking-in region, while the base pressure becomes low and attains minimum.

View full abstract

As a part of the laboratory experiment on an isolated microburst, the instantaneous velocity fields of an impulse-driven vortex ring impinging onto a normal wall were investigated. Vortex rings were visualized by filling ethylene glycol smoke. In this paper, as the sequel of the previous paper (Nagata et al., 2000), by employing PIV, the vertical cross-section of the velocity fields was reported. It was shown that the central downward flow as well as the entrainment above the vortex ring agrees qualitatively with numerical simulation of atmospheric microburst (Proctor, 1984). The roles of vortex ring in the velocity field of microburst was discussed. The time evolution of the total circulation in vertical cross section could be fitted to the theoretical equation for that of incompressible viscous vortex.

View full abstract

Although extended studies have been made on the vortex breakdown inside a cylindrical container by means of numerical simulation, majority of the studies conducted on this flow treated the case of flows inside a closed container with bottom and top disks. The occurrence of vortex breakdown in a cylindrical container is considered to be influenced by the swirling velocity component and the upstream vorticity distribution. Therefore, in the present study, by changing the top disk non-slip boundary condition to the free-slip boundary condition, the influence of the free surface on the emergence of vortex breakdown is investigated in detail. Numerical results are compared with the existing experimental study and novel flow behavior is noticed for container with small aspect ratios.

View full abstract

This study is concerned with the temporal and spatial structure of a radial jet which reattaches on a side wall. Large-scale unsteadiness of the flow was clarified experimentally by means of cross-correlation measurements and conditional sampling technique. The flow oscillates up and down periodically in the arbitrary section and the instantaneous reattachment points vibrate in the reattaching region corresponding to this oscillation. Also, the period increases as the offset distance between the nozzle and the side wall increases.

View full abstract

A direct numerical simulation (DNS) of turbulent free surface flow with a constant magnetic field has been carried out to grasp and understand the effects of electromagnetic suppression of turbulence. In this study, the Reynolds number based on bulk velocity and a film depth was set to be constant, Re_b=2300. The constant static magnetic field at Ha=20 and 30 for the streamwise orientation and at Ha=5 and 10 for the spanwise orientation were considered in the electrical potential equation. The number of computational grids used in this study was 256×128×128 in the x^-, y^- and z^-directions, respectively. The turbulent quantities such as the mean flow, turbulent stresses and the turbulent statistics were obtained. The mean velocity and turbulent intensities distributed from the wall to the free surface are damped quickly near wall region, because of the electromagnetic effects.

View full abstract

In the present study, we construct a turbulence model based on a low-Reynolds-number nonlinear k-ε model for turbulent flows in a rotating channel. Two-equation models, in particular the non-linear k-ε model, are very effective for solving various flow problems encountered in engineering applications. In channel flows with rotation, however, the explicit effects of rotation only appear in the Reynolds stress components. The exact equations for k and ε do not have any explicit terms concerned with the rotating effects. Moreover, a Coriolis force vanishes in the momentum equation for a fully developed channel flow with spanwise rotation. Consequently, in order to predict rotating channel flows, after proper revision the Reynolds stress equation model (RSM) or the non-linear eddy viscosity model (NLEVM) should be used. In this study, we improve the non-linea k-ε model so as to predict rotating channel flows. In the modelling, the wall-limiting behaviour of turbulence is also considered. First, we evaluated the non-linear k-ε model using the direct numerical simulation (DNS) database for a fully developed rotating turbulent channel flow. Next, we assessed the non-linear k-ε model at various rotation numbers. Finally, based on these assessments, we reconstruct the non-linear k-ε model to calculate rotating shear flows.

View full abstract

This paper presents a velocity profile measurement using Ultrasonic Velocity Profile (UVP) measuring method. This method has many advantages over the conventional flow measurement methods. First, it has the capability to measurement of an instantaneous velocity profile on a line. Secondly, it is applicable to opaque liquids. This method can be applied to various flow measurements, but it requires a large measurement volume. In this paper, the ensemble averaged velocity profile and the Reynolds stress have been measured for fully developed turbulent pipe flows in a vertical pipe and the effect of the measurement volume on the measured values was estimated analytically. The results show that the effect of measurement volume appears in buffer region and viscous sublayer.

View full abstract

Direct numerical simulation (DNS) was performed for a spatially developing round jet Re=1200. Streamwise extent of computational domain is 30 nozzle diameter, and initial and nearly developed stages of turbulent jet were predicted. Combination of forth order finite differencing method and numerical damping near the jet center axis resulted in generally good agreement between numerical data and existing literatures with sufficient numerical stability. Detailed discussion was made for Reynolds stress budgets. Inter-component transfer of Reynolds stresses appear both initial and developed stages in a way similar to other one-dimensional shear flows homogeneous in streamwise direction. There were no qualitative differences among stress components for their spatial transport and dissipation. For three components of normal stresses, initially generated turbulence was mostly convected downstream to the developed stage of jet but transport process in the developed stage was dissipative with high conversion rate of kinetic energy into heat.

View full abstract

The logarithmic velocity region is studied in zero-pressure-gradient turbulent boundary layers on the smooth wall. We have already proposed the new definition of log-law region with using the probability distribution function in the stream-wise velocity fluctuation. In this report the Reynolds number dependence of log-law region and PDF profile is discussed based on the experimental data. The measure called Kullback-Leibler divergence is applied for distinguishing the probability profiles. And the mean velocity profile around the edge of log-law region is derived from the PDF equation, which is slightly depending on Reynolds number. The PDF profiles are characterized here by the method adopting the conditional average of the stationary stochastic process. This enables us to understand the important physical quantities which have the relation to the shape of PDF in the log-law region.

View full abstract

A new model for pathological vocal fold vibration is proposed. The model simulates asymmetries in mechanical and geometrical properties of vocal fold, which are pathological features in laryngeal diseases. In this study the effect of asymmetry in vocal fold thickness (L_g) on phonation threshold pressure (PTP) of clinical importance is numerically investigated using the model. The results show that the difference between PTP values in an asymmetric case and those in a symmetric case is small over a wide range of L_g. In the asymmetric case the one of smaller L_g between a pair of vocal folds is found more influential to PTP, and clinical significance of PTP-L_g curve is discussed in relation to surgical treatment for voice disorders caused by unilateral vocal fold paralysis. In the symmetric case PTP-L_g curve is characterised by two distinct regions. Theoretical consideration of the curve from the viewpoint of vocal fold vibration reveals an important role of propagating wave velocity in the region of larger L_g.

View full abstract

Pathological vocal fold vibration is numerically simulated and analyzed using an asymmetric distributed-parameter model, previously proposed by the authors. In the simulations, mass, stiffness, initial glottal half-width at the downstream end and thickness of unilateral vocal fold are chosen as the parameters representative of unilateral vocal fold paralysis, and numerical calculations are carried out for four sets of parameters including one symmetric case. The results show a variety of vibratory patterns such as period-doubling and asymmetries in amplitude and phase, which are comparable with those of patients with paralyzed vocal folds and those observed in model experiments with excised larynges. These dynamical behaviours of great complexity strongly influence speech sound waveforms, and their effects on voice quality are discussed based on spectrum analysis. Furthermore, diversity of vibratory patterns shown by asymmetric vocal folds could be attributable to imbalance in bilateral natural frequencies, made by those parameters.

View full abstract

This paper describes impinging heat transfer from buoyant plume to horizontal downward surface. Experiments are carrried out for a plume produced by heating up the center of the bottom surface settled in parallel to the upper downward surface. The size of heated surface diameter on the bottom wall is D=100 and 300 mm and heat flux of heated surface is changed as q_b=140∼320 W/m². The distance between the two surfaces was changed as H=100∼500 mm and Rayleigh number based on this distance is changed as Ra_H=1.45×10⁶∼4.06×10⁸. The temperature at the stagnation point of the plume colliding on the top surface which is kept at the adiabatic condition is measured and defined as the adiabatic wall temperature. Then the top surace is heated and the surface temperature at the stagnation point is measured again under the same plume impinging condition. The heat transfer coefficient is defined as the difference of these two surface temperatures. The obtained heat transfer coefficients are well correlated by Nu_t0∝Ra¹/3_H.

View full abstract

This paper describes impinging heat transfer from buoyant plume to horizontal downward surface. Experiments are carried out for a plume produced by heating up the center of the bottom surface settled in parallel to the upper downward surface. The size of heated surface diameter on the bottom wall is D=100 mm and terperaure of heated surface is changed as ΔT_bc=16.0∼40.1 K. The distance between the two surfaces was canged as H=100∼280 mm and Rayleigh number based on this distance is changed as Ra_H=1.68×10⁶∼7.93×10⁷. After acceleration is finished, velocity of thermal plume formed from the bottom heated surface approaches to a constant value. It is concluded that the heat transfer coefficient of the top wall at stagnation point becomes a constant as the result.

View full abstract

A method to predict grain structure defects in directional solidification precise casting has been developed based on a solidification process simulation with a three-dimensional CAFE (cellular automaton-finite element) model. In our paper we studied which parameters of the dendrite growth model in the CAFE model are important by sensitive analyses. Then, the dendrite growth model was advanced by evaluating quantitatively liquid diffusion coefficients dependent on temperature and liquidus slope dependent on liquid concentrations which have much effect on the growth model using molecular dynamics simulations and Thermo-Calc calculations. The method to forecast grain structure defects has been applied to directional solidification casting of a simple shape and a gas turbine blade, and has been verified with comparison between the simulated results and experimental results on locations or grain directions of defects, solidification microstructure. It was found that our method could predict a stray grain defect in directional solidification precise casting.

View full abstract

By using the molecular dynamics method, highly accelerated cluster impact on a diamond surface was simulated in order to investigate the surface erosion process. Accelerated argon or CO₂ clusters (∼960 atoms. 100 keV/cluster) impacted on the (111) surface of diamond which consisted of more than 1 000 000 carbon atoms. A typical hemispherical crater appeared about 0.7 ps after the impact, and two or three-layered shockwaves were formed and propagated to certain directions, but the crater was immediately filled up with the fluidized hot carbon atoms due to the collective elastic recovery before the reflection of the shockwave. The impact energy of the cluster was at first transferred mainly as kinetic energy of the diamond surface in a short time, and the potential energy was activated later. The activated carbon and oxygen atoms from the impact cluster stimulated the evaporation from the diamond surface for the CO₂ cluster impact while the evaporation seemed to be suppressed by the argon atoms themselves for the argon cluser impact.

View full abstract

Flow visualization in reciprocating flow inside a three-generation branching-tube network was performed to investigate the mechanism of axial gas exchange in conducting airways of human lung system. Water was used as the working fluid and nondiffusing dye was employed as tracer. The ranges of Reynolds and Womersley numbers at the 0-th generation tube were 60-6000 and 5-25, respectively. It was revealed that bi-directional axial gas-exchange phenomenon occurring in the conductive airways of human lung system was governed by "trap-and-release" mechanism caused by the formation-and-destruction of the separation regions. In inspiration process, separation regions were formed on the outside wall of the two daughter tubes, while they were generated on both (1) the outside wall of the parent tube and (2) the intersection circumference of the two different diameter tubes, or parent and daughter tubes.

View full abstract

Effect of high frequency magnetic field on the Marangoni convection in the floating zone (FZ) silicon melt has been investigated numerically. The purpose of the study is to clarify the applicability of high frequency magnetic field on the suppression of the Marangoni convection in the FZ silicon melt for the space experiments. The numerically obtained results reveal that the high frequency magnetic field has the damping effect for the Marangoni convection, and is useful to control the melt convection in the space environment. Further, the higher magnetic field frequency is useful to suppress the Marangoni convection due to the skin effect. The impurity distribution in the crystal is also affected by the high frequency magnetic field and the impurity distribution respect to the radial direction becomes rather flat at the mostly suppressed Marangoni convection condition.

View full abstract

An experimental study has been made to elucidate an effect of oscillated flow induced near critical heat flux and natural circulated flow of vapor and liquid in a vertical tube on the critical heat flux. The experiment has been carried out for the condition of heated tube length of L=100 to 1000 mm, tube diameter of D=4 to 9 mm and the CHF was measured under the condition that the exit of the unheated tube connecting the heated tube was extruded into a vapor chamber to prevent a liquid flowing into the heated tube from the top of the tube. The experiment reveals that the CHF for D=7 and 9 mm and any tube length from 100 to 1000 mm becomes identical to that for the case of the tube top being in the liquid, while the CHF for D=4 and 5 mm is smaller than that for the case of the tube top being in the liquid due to no liquid supply from the top. It is found that the oscillation induced near the CHF increases the CHF.

View full abstract

The vortex formation and the turbulence-energy production in a wall-impinging jet were investigated by the instantaneous and time-averaged velocity distributions measured using cross-correlation of high-speed particle images. By comparing the measured results with those in a free jet, the characteristics of flow patterns around the wall were clarified. Based on the ensemble of fluctuating and instantaneous velocities, the production term of the budget equation of turbulence-energy k was directly evaluated. The rate of k-production increases where flow was decelerated due to the normal stress to the wall. In addition, some specific vortex-patterns were sampled, which exemplified an increase in the rate of k-production around the stagnation point.

View full abstract

A laser diode absorption method to measure CO₂ gas temperature and concentration was developed using a 2.0 μm DFB (Distributed Feed Back) laser. Optimized two absorption lines for CO₂ around 5000 cm^-1(2.0 μm) were selected from HITRAN 98/HITEMP database. Two sets of 2.0 μm DFB laser were fabricated in a pigtail fashion and all components were contained in a box. Special optical isolator, a single mode fiber for 2.0 μm, a fiber ring interferometer and optical filter for wavelength separator were all developed for the system. The measurement sensitivity at different sweep frequencies from 100 Hz to 10 kHz was evaluated using heated gas cell which are contained CO₂ gas. The results sbowed that the system has a small error under high concentration and low temperature. Accuracy, in low concentration and high temperature show some tolerance (3.64%) because of small absorption. The comparison between HITRAN 98/HITEMP database and our experimental data were performed to establish our data source for temperature/concentration measurement.

View full abstract

We developed in the previous report a simple method of measuring thermophysical properties of ceramic materials by stepwise heating. One of the important points in the method is the measuring technique of the surface temperature of a specimen in a simple and low-cost manner: the surface temperature is measured by pressing separated thermocouple wires against the specimen surface to which an aluminum foil is attached to provide high electric and thermal conductivities. In this repot the effects of the specimen thickness, heating time and the thickness of the thermocouple wire and the aluminum foil on the accuracy of the measurement have been theoretically investigated for thirteen materials, and the optimum measuring condition of the preset method has been clarified.

View full abstract

Pressure distributions of dense arrays of in-line circular cylinders, imitating arrays of condensers on the wall of parallel channel were measured. The diameter and height, d and H, of the cylinder were 16 mm and 20 mm, respectively. The effects of the cross-stream and stream-wise pitches of the cylinders, P₁, P₂, the number of row, N, and the other factors on the pressure loss were clarified. From the flow visualization using oil-film method, the flow pattern is divided into two regimes: one is fence-type for P₁/d≤1.2 and the other is In-line-type for P₁/d>1.2. And the pressure drop is divided into three parts: the inlet, intermediate and outlet ports of the arrays of the cylinders. Depending on the above flow patterns, these pressure drop and pressure recovery are formulated by the dominating factor of the opening ratio, β, of the channel, the pitch, P₂, and numbers of row, N, of the cylinders. The pressure loss given by the recommended equation agrees with the experimental data within ± 10%.

View full abstract

Micro-gravity experiment was conducted to observe the instability of a SF₆ liquid jet issued into an otherwise quiescent N₂ gas at high pressures exceeding the SF₆ critical pressure. The liquid, which had a near-critical mixing surface, had an extremely small surface tension but a relatively large disparity in liquid-and gas-phase densities. The disintegration process of O(1)-Weber-number liquid jet was successfully observed to reveal the importance of fluid dynamic action in atomization which has been overlooked in past studies. An instability wave which had a wavelength comparable to the jet diameter developed. Droplets disintegrated from the liquid jet made pairs to coalesce, so that the droplet spacing increased downstream.

View full abstract

The experimental results presented in the first report are physically interpreted on theoretical grounds to reveal the disintegration mechanism of the O(1)-Weber-number liquid jet with nearcritical mixing surface. The wavelength at which the liquid jet breaks up is much shorter than that of Rayleigh instability, suggesting that the shear layer developed on the near-critical mixing surface plays an important role in excitation of the instability. Vortex rings, which grows by Kelvin instability, entrain and impinge the surrounding dense gas on the contracting part of the liquid column, so that the restoring force associated with the axial surface tension is cancelled by the dynamic pressure of the impinging gas, and the liquid column breaks up by the action of contracting force due to the circumferential surface tension. This mechanism leads to a concept which bridges the Rayleigh type of atomization and the large-scale vortical structures of a supercritical liquid jet.

View full abstract

Economic viability on using a polymer electrolyte fuel cell (PEFC) as a cogeneration unit for household application is investigated. The PEFC cogeneration unit is installed with a thermal storage tank. An optimization approach is adopted to rationally evaluate the viability. Design variables such as equipment capacities are determined together with the system's operational strategies so as to minimize the annual total cost subject to satisfy all the possible energy demands. This optimization problem is formulated as a large scale mixed-integer linear programming problem by using appropriate approximation. Numerical studies are carried out with some parameters such as energy efficiency and capital unit cost of the PEFC cogeneration unit, unit cost of natural gas energy charge and load factor of household energy demands. Through the studies, it is found that the relationships between the ecomomic viability of PEFC cogeneration system and the aforementioned parameters are clarified.

View full abstract

Particulate matters (PM) emitted from Diesel engine vehicle have caused the serious social problem (e.g. carcinogenic to the human body). The additional energy is necessary in most of the filter/combustion devices of particulate matters developed recently, which do not essentially constitute the self-sustained combustion system. While the self-sustained combustion has been developed in the super-adiabatic combustion with reciprocating flow in a porous medium, the feasibility of the PM combustion is examined based on the analytical model taking account of the transient behavior for intermittent filtering and/or combustion processes. The validity and of preceding concept, being supported by the preliminary experiment and the prominent features compared with the conventional devices, has been demonstrated.

View full abstract

Rapid mixing of fuel and air is an essential factor in improving combustion and emissions of diesel engines. The paper investigates local diffusion phenomena with focusing on scales of fuel cloud and eddies. To analyze degree of heterogeneity entropic analysis was developed and applied to the diffusion phenomena in turbulent jets and diesel sprays. The result shows the effectiveness of the entropic analysis method for identification of scales of heterogeneity and diffusion intensity. The result of the experiment shows that the diffusion intensity is the highest at the vicinity of the nozzle exit, and size scale of the heterogeneity is the smallest. The scale of the heterogeneity increases gradually along the spray axis towards downstream, with containing smaller scales inside of the large clouds. In the downstream region, small-scale structure diffuses and becomes unclear, while large scale structure remains clearly. The paper reveals microscopic structure of the heterogeneity in diesel sprays, and it demonstrates availability of the entropic method.

View full abstract

The local reaction rate and the diffusion rate in a turbulent diffusion combustion field are important parameters which dominate combustion phenomena in a turbulent diffusion flame. However, they can not be directly measured. Then, the present authors developed the method to obtain the local reaction rate and the diffusion rate by numerical calculation using measured values. In the present study, this method was applied to a CO/H₂ jet diffusion flame in order to examine the influence of preferential diffusion of H₂ on a jet dilusion flame. The following results were obtained. (1) In a high-temperature region, the effective diffusion coefficient D_H, which is obtained using H element as tracers, was larger than the effective diffusion coefficient Dc which obtained using C element as tracers. This result indicates that the molecular diffusion rate of H₂ increases in a high-temperature, and that it can not be disregarded in comparison with the turbulent diffusion rate. (2) The large difference in the reaction rate distribution was produced by the difference of the molecular diffusion rate between H₂ and CO. The local reaction rate of H₂ was more than 2 times in comparison with that of CO. The peak position of the reaction rate of H₂ was remarkably more located in the fuel side than that of CO.

View full abstract

With a mind to develop a model capable of dealing with the turbulent combustion with the local extinction, flame stuctures with the local extinction have been numerically investigated in the counterflow configuration. The mass, momentum, species, and energy conservation equations were solved by the SIMPLE algorithm to seek steady state solutions. A Lewis number of unity and a one-step irreversible reaction with a finite rate were assumed. To describe flame tip structure, a reaction progress variable based on the amount of a product is introduced in addition to the mixture fraction and the scalar dissipation rate. It is found that the flame tip structure is divided into three categories; say, the fully burnt flamelet, the intermediate flamelet, and the preheat region and is determined by the above three parameters at an accuracy level. Especially, the first two of flame structures are uniquely determined by the three parameters; however, the last one has revealed a small deviation attributable to the diffusion process. However, so long as the scalar structure of the preheat region is kept to be similar to the flame structure, the three parameters are important factors determining the flame tip structure.

View full abstract

Structure and characteristics of the plane shear layers in the two-dimensional divergent or convergent duct are numerically examined by using the discrete vortex method combined with the Schwars-Cristoffel transformation. The transformation of such complicated flow configurations is successfully accomplished by introducing an original conformal mapping function. The growth rate of the shear layer is also analyzed by estimating the momentum thickness of the shear layer. It is shown that the shear layer, which is composed of a series of clusters of discrete vortices, shifts from the lower to the higher mainstream region with increasing the divergent angle of the ducts, resulting in a remarkable reduction in the lower mainstream velocity. Increasing the divergent angle of the duct increases the scale of organized eddies and therefore promotes the growth of the shear layer, whereas increasing the velocity ratio reduces the scale of organized eddies and therefore suppresses the development of the shear layer. The characteristics of the shear layer calculated show good agreement with experimental results, indicating reasonable accuracy of this numerical simulation.

View full abstract

Forced oscillating combustion can reduce nitrogen oxides (NO_x) as described by the previous research. In this paper, we applied this method to two kinds of radiant tube burner in order to confirm the effect of NO_x reduction. Fuel is supplied through a solenoid valve which is made open and shat several times a second with various duty ratios. Flow rate of the fuel, temperature of the furnace and air ratio are switched to investigate the effect. About 30% of NO_x reduction is observed under acceptable performance of combustion and this method is proved to be applicable to many combustion appliances.

View full abstract

The location of knock initiation was analyzed using the random motion simulation in a combustion process in an s.i. engine. The spatially averaged tranport equations were solved in conjunction with a reduced kinetic model for the low temperature oxidation and with a combustion model for a turbulent flame propagation in a cell-averaged form. The prediction for a central spark case shows that the computed time of the first local auto-ignition agrees well with an experiment. The simulation for an off-center spark case shows that the first autoignition sites are in the locations comparatively close to the wall and not in the core region of the unburned mixture.

View full abstract

From an ecological point of view, vegetable oils are a renewable, biodegradable and nontoxic fuel with the potential to reduce the emissions of CO₂ Palm oil is one of effective vegetable oils as alternative diesel fuel because of the high productivity and low cost. In this study, to utilize palm oil methyl ester (PME) as a diesel fuel, the engine performance with PME and the exhaust emissions from PME are investigated using a DI diesel engine. In addition, the investigation is carried out in comparison with rapeseed oil methyl ester (RME) and gas oil. The results show that ignition delay of PME is shorter than those of RME and gas oil, that brake thermal efficiency with PME is about the same as one with gas oil, and that NO_x and Smoke emissions from PME simultaneously are lower than those from gas oil. Therefore, PME can be used as an alternative fuel of petroleum diesel fuels.

View full abstract