The interfering models are represented to decide the rate of the dividing or combining at the junction of the combination pipe. In the conventional theory of the energy loss, it is all but impossible to determine the flow ratio of any ramifications, because the energy loss is a result of distribution of flux of flow and is not a cause of it. In the interfering model, the rate and the pressure distribution at the junction are determined by using coefficcients depending only on the geometrical scheme of the junction. Coefficients of interference are determined by the fitting method using the experimental data of the conventional energy loss coefficients. Rate calculated by using the interfering model are in accord with the well-known experimental results. Through the consideration about the case of simple network, it is shown that the rate calculated by the model coincide with the result of the Hardy-Cross method in the case of long length piping.

View full abstract

A prototype system for feedback control of wall turbulence is developed. and its performance is evaluated in a physical experiment. Arrayed micro hot film sensors with a spanwise spacing of 1 mm are employed for the measurement of the streamwise shear stress fluctuations, while arrayed magnetic actuators of 2.4 mm in spanwise width are used to introduce control input through wall deformation. A digital signal processor having a time delay of 0.1 ms is employed to drive output voltage for the actuators. Feedback control experiments are made in a turbulent air channel flow. A noise tolerant genetic algorithm is employed to optimize control parameters. It is found that the wall shear stress is decreased by about 6% experimentally for the first time. The Reynolds shear stress close to the wall is decreased by the present control. By using conditional average of a DNS database, it is demonstrated that wall induces wall normal velocity away from the wall, when high speed region is located above the actuator.

View full abstract

Turbulent concentric annular pipe flows with several thin inner rods were invistigated by means of the direct numerical simulation (DNS). In the inner region, the mean velocity shifts downward from the logarithmic law and the Reynolds stresses normalized by the inner-wall friction velocity decrease monotonously as the radius ratio decreases. In the budget of the Reynolds stress for the very small radius ratios, the pressure-strain term is restrained near the inner wall. We found that the skewness, flatness, two-point correlation and instantaneous field for very small radius ratios are much different from those for large radius ratios.

View full abstract

Experimental studies on mass transfer through the gas-liquid interface in a micro channel were carried out to investigate gas absorption rates against flow patterns and the efficiency of gas absorption. The micro channel used was a glass capillary tube which had a diameter of 0.5 mm and carbon dioxide was absorbed into distilled water in the micro channel. The superficial velocities of gas and liquid were widely changed to form various flow patterns in the channel. The results show that the gas absorption is remarkably progressed as the gas flow rate increases. The volumetric mass transfer coefficient, k_La, generally has a linear relationship with gas superficial velocity, regardless of the liquid flow rate and flow patterns. k_La obtained in the micro channel is much larger than those in the conventional methods.

View full abstract

The nonlinear time series analysis of two-phase flow is difficult, and the analysis sometimes results subjectively. Using artificially generated quasiperiodic data and submerged nozzle pressure fluctuation, we showed that applying an optimal delay time reconstruction method and a point correlation dimension method together greatly improved the reliability of the analysis. In this paper, we analyze real two-phase flows time series : differential pressure and void fraction of slug and bubbly flow in a vertical pipe. The comparisons of estimated results and their statistical test results demonstrate the high reliability of the proposed analysis method.

View full abstract

Prandtl's secondary mean motions of the second kind near an undulating surface were explained in terms of turbulent blocking effect and kinematic boundary conditions at the surface, and its order of magnitude was estimated. Isotropic turbulence is distorted by the undulating surface of wavelength λ and amplitude h with a low slope, so that h<<λ. The prime mechanism for generating the mean flow is that the far-field isotropic turbulence is distorted by the non-local blocking effect of the surface to become anisotropic axisymmetric turbulence near the surface with principal axis that is not aligned with the local curvature of the undulation. Then the local analysis can be applied and the mechanism is similar to the mean flow generation mechanism for homogeneous axisymmetric turbulence over a planer surface, i.e. gradients of the Reynolds stress caused by the turbulent blocking effect generate the mean motions. The results from this simple analysis are consistent with previous exact analysis in which the effects of curvature are strictly taken into account. The results also qualitatively agree with flow visualization over an undulating surface in a mixing-box.

View full abstract

This paper deals with the flow around a circular cylinder with an element inserted in the wake. A long splitter plate (splitter-plate case) and a circular cylinder of equal diameter (two-cylinder case) were used as the element. Experiments were mainly conducted in a water tank at a Reynolds number of 7.4×10³ for the two-cylinder case and 1.1×10⁴ for the splitter-plate case. As the deflectors are horizontally traversed upstream, the base suction coefficient exhibits a critcal tall at a certain position. PIV measurements revealed suppression of regular vortex shedding at the critical positions in both cases and a similar behavior of vortex formation length in the vicinity of the positions.

View full abstract

When a high-speed train enters or leaves a tunnel, impulsive pressure waves are radiated from the entrance/exit portal toward the outside of the tunnel. The waves are called a “micro-pressure wave, ” a “tunnel entry wave” and a “tunnel exit wave.” A similar phenomenon occurs when a train passes a short structure in axial length in an open section (such as an over-bridge). We call it a “structure-passing wave.” The length of the structure along the railway is several times or less the diameter of the structure cross-section. Although the structure-passing wave is weak at present, it could cause an environmental noise problem of a low-frequency (which has become public concern) as the train speed increases. This paper describes the characteristics of the structure-passing wave based on the results of model experiments and acoustic analyses. The results obtained from the acoustic analyses were in good agreement with the result of the experiment except for the very short length structures. This implies that the structure-passing wave can be modeled on a superposition of the pressure waves radiating from both sides of the structure.

View full abstract

Cloud cavitation is known well as one of the most destructive cavitation. High pressure generated by the violent collapse of a bubble cloud causes severe erosion, noise and vibration in hydraulic machines. However, this high concentration of the energy has been utilized recently for biomedical, environmental and other industrial applications. In the medical field, the behavior of the bubble cloud has much influence on HIFU (High Intensity Focused Ultrasound) applications, where acoustic cavitation is generated due to the high pressure amplitude in the focal region. In this study, it is assumed that the acoustic cavitation forms a spherical bubble cloud which consists of many microbubbles, and ultrasound focusing in the spherically symmetric cloud is numerically investigated. We consider the compressibility of the liquid, the evaporation and condensation of the liquid at the bubble wall, heat transfer through the bubble wall in the simulation. The pressure wave focuses to the center of the cloud and the pressure inside bubbles extremely increases when the frequency is near the first mode natural frequency of the cloud. Especially, in the cases of relatively high pressure amplitudes, a shock wave is formed in the cloud and it generates high pressure fluctuation near the center of the cloud even when the frequency is much lower than the first mode frequency.

View full abstract

Spectroscopic measurements of micro-air plasma-jets generated at low pressures were performed at several points on flow axis, using a spectrometer in the wavelength region of 200 to 900 nm. The N₂2+ bands and N₂⁺1-bands were predominant in the region of 280 to 480 nm, and atomic lines of N and O in the visible and infrared regions. Temperatures of the plasma-jets were estimated by applying a spectral matching method to N₂⁺1-bands along a flow axis for low-pressure conditions. Rotational temperatures for N₂⁺1 bands were 6000 to 7000 K, and vibrational temperatures 25000 to 28000 K, almost independent of pressures and axial positions. It was found that the rotational temperatures measured had great difference from the gas temperature based upon an empirical formula for supersonic free jets, and also the vibrational temperatures were much higher than those expected.

View full abstract

It is well known that a steam control valve may cause severe flow vibration at partial valve opening, in thermal or nuclear power plants. For the rationalization of maintenance and management of the plant, the vibration should be avoided. However, it is difficult to understand the flow characteristics in detail only by experiments because the flow around the valve would he very complicated with 3 D supersonic structure. Under this circumstances, the application of the latest CFD (Computational Fluid Dynamics) technology is useful for the clarification of the cause of the vibration and the examination of the methods for suppression. In our previous research, it was found by combining experiments and CFD that the interaction of the flow attached to the valve body causes a pseudo-periodic spike type pressure fluctuation at partial valve opening. Since this pressure fluctuation results in cyclic side load on the valve body, it will cause lateral vibration of the valve body. In the present report, several geometries are proposed and examined for the suppression of the flow oscillation at the partial valve opening, based on the understanding of the mechanisms of the flow oscillation as mentioned above. First, the change of the valve seat geometry such that the flow is enhanced to attach to the valve seat was examined. Although some effects could be found, we could not succeed in completely suppressing the oscillation. Then, two types of valve body geometry, which would prevent the flow attachment to the valve body are proposed and examined. It was found that one of them could completely avoid the flow oscillation.

View full abstract

Ultrasonic Velocity profiler (UVP) is a powerful tool to measure an instantaneous velocity profile especially on a velocity measurement of an opaque liquid flow, such as liquid metal. In this study, UVP is applied to a velocity field measurement of a fluid flow of Lead Bismuth Eutectic (LBE) in the Accelerator Drive System (ADS). At a lower temperature, wetting of LBE to stainless steel that is a material of target loop is too poor. We solved this problem by solder coating of the measurement window of the target loop. We performed velocity measurement on the centerline of the loop and confirmed basic performance of the loop. It was found that there was intermittent release of eddy from the re-circulation region formed on a wall surface of an inner cylinder and behind the edge. We made then a measurement for an angluar direction of circumference around the central axis and obserbed 3-dimensional structure of LBE flow of the loop. From these results, we made clear that UVP is highly effective in a measurement of the flow field of liquid metal.

View full abstract

Flow visualization with surface tufts is carried out to investigate the flow behaviors in two dimensional separated flows over a flat plate. The separation is realized in the channels with diverging cross sectional areas, which are designed to generate constant adversed pressure gradients with quasi one dimensional flow assumption. Distributions of the local flow direction on the flat plate are determined from the photographs of the flow visualization with surface tufts. The flow angles are presented with complex numbers to enable ones to treat them in various manners. Complex correlation factor is introduced to evaluate the three dimensional behaviors of separated flows. The results show the flows immediately downstream of the separation exhibit periodic correlative regions in the spanwise direction as well as in the downstream direction, and their scales are independent of the flow velocity and the magnitude of pressure gradients.

View full abstract

As a purpose of developing micro-actuators driven by liquid crystals, unsteady behaviors of a nematic liquid crystal between two parallel plates under an electric field are investigated both numerically and experimentally. Imposition of the electric field on the liquid crystal induces flows, whose profile and magnitude depend strongly on the twist angle of the director ; that is, the velocity profile between plates is S-shaped and the maximum velocity is about 80 μm/s when the twist angle is 0 deg. On the other hand, unidirectional flow is generated and the maximum velocity is 350 μm/s when the twist angle is 180 deg. Generation of flow is confirmed by means of a flow visualization experiment. Mechanism of the generation of flow can be explained by the consideration that the rotation of liquid crystalline molecules generated by imposition of an electric field induces local velocity gradient.

View full abstract

An experimental study by means of pressure measurements and flow visualization was performed to investigate unsteady flows inside a two-dimensional semi-open-type nozzle for a ship propulsion equipment directly driven by high-pressure gas. We found that ejected gas phase and water-flow phase are separated clearly of themselves, and the interface of these phase behave like interfacial waves. It is clarified by flow visualization with a high speed motion camera and a circulating water channel that these interfacial waves change their shapes according to water-flow velocity. The interfacial wavelength becomes longer in response to increasing water-flow velocity, and the mechanism that obtains thrust on the nozzle-wall changes. The thrust and flow patterns for the intermittent gas ejection according to water-flow velocity are also clarified.

View full abstract

Small axial fan which diameter is less than 100 mm has been used widely as cooling device. Experimental studies have been carried out in order to improve performance and reduce noise of small axial flow fan. In this report, small axial flow fan with diameter of 92 mm popularly used at a personal computer or a workstation was chosen, influence of design parameters of impeller such as camber of blade, maximum camber point of blade, cross sectional figure of blade, skew angle of blade and roundness of blade edge on aerodynamic performance and noise were investigated.

View full abstract

The turbine used for a wave-power-generating system is driven by the oscillating airflow produced by the water wave motion. The Wells turbine has the simple vane less structure using no-twist and symmetrical blades, which drives its unidirectional rotational motion even though the turbine is driven in the oscillating airflow. In this paper. the important characteristics of the turbine are theoretically derived by the potential theory of cascade. The deflection angle in the absolute field does not depend on any angle of attack, but is always constant. The deflection angle depends only on the arrangement of the cascade. Moreover, the pressure drop coefficient is proportional to the rotational speed at the same axial velocity and proportional to the axial velocity at the same rotational speed.

View full abstract

To cope with the warming global environments, greenhouse gas emission should be reduced in the electric power generation. The hydroelectric power should occupy the attention of the electric power generation systems, as clean and cool energy sources with the highest density. The authors have proposed the counter-rotating type hydroelectric unit, and discussed the fundamental performances and the flow conditions. This paper discusses the effects of the blade profiles on the hydraulic performances. As a result, the design materials for the solidity of the axial flow runners suitable for the given water circumstances are induced from above discussions.

View full abstract

Two-dimensional numerical computation has been performed for unsteady laminar flow. Spatial periodic boundary conditions were adopted in the streamwise direction and adequacy of its treatment was particularly studied. The following results were obtained. When the numerical result fluctuates with the computational domain size, various low frequency components are observed remarkably in the streamwise velocity of the calculated result. Compared with channel flow obstructed by single rod of the same size, intensities of velocity fluctuation in this study are relatively high, and their distribution throughout the flow field is rather uniform. Averaged kinematic energy of the velocity fluctuation is adopted as a key parameter, and a discrimination map of the validity of numerical results is presented with a time scale ratio between mean flow and velocity fluctuation, the rod pitch and the blockage ratio. The time scale ratio can be well correlated with a geometric parameter obtained from the flow configuration, and the discrimination map can be converted to a map described with the geometric parameter and the channel Reynolds number.

View full abstract

The fuel assembly design with tightened fuel rod pitch gives a possibility of higher conversion ratio for Boiling Water Reactor. Pressure drop is one of the key design parameters to evaluate the possibility of BWR core with higher conversion ratio. The pressure drop tests for tight lattice rod bundle under steady conditions have been done under BWR fluid conditions. Two kind of a mini tight lattice bundle were used. One was a 7-rod test bundle with a hexagonal channel box. The other was a 14-rod test bundle with a rectangular channel box. The purpose of the 7-rod bundle tests was to survey the rod gap effect and the heating length effect on pressure drop. That of the 14-rod bundle test is to check the bundle shape and flow shape. The each pressure drop elements (head loss, friction loss, acceleration loss and spacer loss) were predicted based on the appropriate correlations for each item. The total pressure drop was obtained by sum of each pressure drop element. It was found from the compassion with data that the present prediction method has reasonable performance. The average and the standard deviation for the ratio of the predicted and the measured value is 1.05 and 9.7%.

View full abstract

The purpose of present study is to develop the Soret forced Rayleigh scattering method for measurement of mass diffusion coefficient. This method has attractive characteristics such as (1) short measurement time of 100 ms, (2) small sample volume less than a few μl, (3) the concentration dependence of diffusion coefficient is detectable and (4) the applicability to measurement of the anisotropic diffusion coefficient. In the present paper, the verification of the measurement system and the measurement of concentration dependence of the mass diffusion coefficient in Fullerene solution are discussed. Measurement results revealed the concentration dependence of the diffusion coefficient of fullerene/o-dichlorobenzene solution.

View full abstract

The Soret forced Rayleigh scattering method (S-FRSM) for measurement of mass diffusion coefficient has been developed. In the present paper, the theoretical analysis of the systematic effects of the experimental setup is discussed. It includes the analysis of : (1) effect of dye, (2) effect of sample thickness, (3) effect of Gaussian beam intensity distribution and (4) effect of heating duration time. The analysis of three-dimensional diffusion equation which takes into account the practical experimental conditions has clarified the influence quantity of the inadequate experimental settings. According to the present analysis, it has been found that the difference of diffusion coefficient can be negligible by employing the appropriate setting of experimental parameters. As a result of these considerations, the reliable optimization method of the measuring conditions in S-FRSM can be provided.

View full abstract

The purpose of this study is to develop effective drying method for seafood. We conducted the warm-air drying experiments using the scallop and investigated the heat and mass transfer mechanism. We have observed morphological changes of muscular fibers of scallop during the drying process. It is found that the surface shrinks due to the rapid drying which results in an additional resistance for water transport and reduces the evaporation rate. This decrease of evaporation rate raises the scallop temperature and increase the water content inside the muscular cell due to osmotic pressure change. This is an additional resistance for mass transfer. A heat and mass transfer model is made by considering these two resistances and we have good agreement with the experiments.

View full abstract

This paper deals with experimental and analytical heat transfer study of nonadecane spheres melting in natural convection of water. Experimental local and average heat transfer coefficients were obtained by analyzing the photo images of the shape change of nonadecane spheres. The relevant dimensionless parameters were varied in the ranges of 1.8×10⁷≤Gr_n·Pr_n≤1.8 × 10⁸ and 0.04≤C_nΔT/L_n≤0.16. We found : (1) Melting nonadecane sphere was covered with thin nonadecane liquid film flowing upwards along the sphere to make a liquid cap on the top. The cap regularly repeated formation and splitting off. (2) The experimental average Nusselt number was correlated by Nu=0.151 (Gr_n·Pr_n) ^0.257 (C_nΔT/L_n) ^0.117 that was in good agreement with analytical one. (3) On the lower hemisphere experimental heat transfer coefficients were slightly higher than analytical ones and on the upper hemisphere the experimental heat transfer coefficients were slightly lower than analytical ones.

View full abstract

An experiment has been made to measure the rise in temperature of a gas during filling a tank at high pressure. The experimental condition is that filling gases are Nitrogen and Hydrogen at a pressure of 5 to 35 MPa and at a filling mass of G=45 to 324 g/min for Hydrogen. The temperatures are measured either horizontally or vertically at five positions in the tank. It is found that heat loss transferred from compressed gas to the tank wall has a significant effect on the rise in the filled gas temperature. The heat transfer coefficient is estimated after the end of filling and is about α_h=270W/ (m²K) for the Hydrogen at 35 MPa. A theoretically procedure is proposed to calculate the temperature increase of the gas on a basis of assumption that the gas temperature in the tank is uniform at any time, and the heat transfer coefficient is given. The calculation shows that the temperature is in reasonable agreement with the measured temperatures by assuming α_h=500 W/ (m²K) during the filling of hydrogen at 35 MPa, although the estimated heat loss after the end of filling becomes larger than the actual one.

View full abstract

Output voltage stability and product water behaviour in cathode gas channel are observed by the visualization of PEFC with dual serpentine gas channel. Result show that output voltage drop starts at the beginning of plug flow and reaches minimum value at the end of it. Drop voltage is in proportion to plug flow locus length. The plug flow is diminished by leakage to the other channel through clearances between by-polar plate and gas diffusion layer, this indicates small bypass channel to the other channel makes plug flow small. Excess saturated water in the piping from humidifier to cathode gas inlet makes plug flow simultaneously in both channels. To avoid emergency stop under limited voltage, it is important to keep less saturated and condensed water at cathode air inlet by heat insulation and by setting water trap.

View full abstract

The performance and over voltages of the polymer electrolyte fuel cell (PEFC) were evaluated using a test apparatus in which the clamp load of a test cell could be changed between 100 N and 1 000 N. The comparison of the performance between paper and cloth type gas diffusion layers (GDL) was also evaluated. The effect of GDL on the PEFC performance varies greatly depending on the clamp load and gas humidification condition. Under moderate humidification condition, an increase in the cell clamp load is effective in reducing IR over voltage, and thereby improving the PEFC performance significantly. However, the high clamp loads over 600 N contribute slightly to the improvement in the performance. The performance of the paper type GDL is higher than that of the cloth type GDL at high clamp loads. Under high humidification condition, however, the cloth type GDL shows lower concentration over voltage, resulting in the more excellent performance at a relatively low clamp load of 300 N, compared with the paper type GDL. However, when the clamp load becomes too high, because the ability to reduce concentration over voltage of the cloth type GDL is degraded, a significant difference in the performance between the paper and the cloth type GDLs can not be recognized.

View full abstract

Effects of design parameters in paper type GDL (Gas Diffusion Layer) are analyzed by experiments, which are hydrophobic treatment by PTFE, MPL (Micro Porous Layer) and carbon matrix density. Results show that hydrophobic treatment with 5% PTFE decreases concentration overpotential, but with 20% PTFE increases it at higher current density region. MPL makes concentration overpotential lower at higher current density region, in spite of the lower Gurley number by 1/80. Moreover, MPL decreases IR overpotential by improving electrical contact between electrode and GDL under lower contact pressure. To increase Gurley number by reducing matrix density makes concentration overpotential lower.

View full abstract

Effects of hydrophobic treatment and MPL (Micro Porous Layer) in PEFC were investigated by experiments at starting test procedure mode. Hydrophobic treatment decreases accumulated water in DGL (Gas Diffusion Layer), and increases it on cathode electrode, which disturbs gas diffusion and makes PEFC output voltage lower. Coating of MPL on GDL protects cathode electrode wetting and improves PEFC performance. Improvement of gas permeability of GDL matrix makes not only gas diffusion to cathode electrode smooth but also exhausting of product water vapor from electrode to gas channel active, and these two merits improve PEFC performance under flooding obstruction at starting procedure mode.

View full abstract

High-resolution particle image velocimetry was used repetitively to obtain the time-change of instantaneous velocity vectors, which enabled us to analyze turbulence kinetics. Each term in the balance equation of turbulence energy was evaluated, based on the ensembles of fluctuating velocities and instantaneous velocity-gradients, thereby finding the factors primarily responsible for production and dissipation of turbulence energy. It was shown that the turbulence production rate is high in the shear layer because of a large Reynolds stress, and viscous dissipation of turbulence energy occurs mainly near the jet axis. Furthermore, typical turbulence eddies in the shear layer were sampled to discuss the effects of instantaneous vortices and fluctuating flow on turbulence production. It was shown that elliptical shape of vortex observed in the shear layer might significantly contribute to the turbulence production. Also, the effect of vortex shape and arrangements on the local turbulence production were discussed in more detail. There were two regions with turbulence production around the vortex and the local turbulence dissipation distributes in the center of the vortex.

View full abstract

As a first step to clarify the mechanisms of the combustion and gasification of biomass, fundamental experiment for the thermal decomposition and numerical simulation for the elementary reaction of decomposed chemical species have been performed. In the experiment, various kinds of biomass particle were immediately heated by applying the electrically induced magnetic field to the metal tray on which the objective particles were placed. It was clarified that for the two levels of decomposition temperature 590 and 1040°C, the variations of decomposition rate with heating time was same with each other, but the components of decomposed species were largely different; the mole fraction of C6-C9 components for the low temperature 590°C were 3-10 times larger than those for 1040°C. Furthermore, the relationship between percentage of C5, C6-C9 species and lignin containing in biomass was ascertained. It was obtained from the numerical simulation that these heavy species could decompose to C1 and H₂ at 1000°C over with H₂O acting as the oxidation agent, and that the effect of H₂O in the atmospheric pressure became maximum at P_H2O-0.7 atm.

View full abstract

Hydrogen is the most ideal and promising alternative fuel, and fuel cell vehicles that use hydrogen as a fuel are expected to be widely used in public. To use hydrogen commonly, the establishment of proper guidelines for handling safely is required urgently. In a series of research to develop safety data of hydrogen systems, the accurate evaluations of the effect of blast wave from hydrogen explosion were examined. In the case of deflagration, measured blast pressure is influenced largely by thermal strain of the diaphragm of the pressure sensor due to the radiation from the flame. In present work, reduction of the effect of radiation to pressure sensor with recess mount was estimated by calculation using view factor. The natural frequency and the response of the recess mount sensor were also evaluated. The validity of recess mounted sensor was shown through the experiments on the blast waves measurements of hydrogen/air denigration. Blast waves were measured by using both pressure sensors with recess mount and with conventional flush mount. The recess mount method suppressed the effect of radiation remarkably without lacking the accuracy.

View full abstract

Turbulent diffusion flame is supposed to be composed of a lot of laminar diffusion flamelets. To understand extinction processes of the flamelet, counterflow diffusion flame has been used. It was well known that the extinction strongly depends on stretch rate. But recently, it became to be known that the extinction limit was influenced by convection flow along the flamelet. Numerical computations of laminar counterflow diffusion flames have been conducted. It has been clarified that there are two types of extinction. One is intermittent extinction, and the other is perfect extinction. Extinction occurs slightly in downstream where local stretch rate is the largest. The maximum local stretch rate is much larger than extinction limit of flat flame. We investigated the effects of covection flow along the flamelet and clarified how much the extinction limit changes.

View full abstract

Evolution of the flame-front movement in burner stabilized turbulent premixed flames in a reaction sheet regime has been examined. At the almost entire position except the vicinity of the burner axis in the turbulent flame brush, flame fronts pass the stationary measuring position alternately in the burnt-to-unburnt then unburnt-to-burnt directions. There is an influence of the direction of flame front passage. When the flame front passes the measuring position in the burnt-to-unburnt direction, it moves away from the burner axis, but when the flame front passes in the unburnt-to-burnt direction, it moves toward the burner axis. In the former direction, the flame-front movement is follow the gas flow, while in the latter direction, the flame-front movement is against the gas flow.

View full abstract

Recent supercomputer advances enable 3-D Direct Numerical Simulation (DNS) of turbulent combustion to be performed. In oral presentations, computer-graphic animations of such results extensively demonstrate the 3-D movement of the flame front and its interaction with turbulent eddies. Only a limited amount of experimental data are available, however, that can be compared with the 3-D DNS results because of great experimental difficulties in measuring 3-D flame-front movements and configurations. It is therefore desirable to establish a definitive technique to measure the 3-D movement and configuration of a flame front in premixed turbulent flames. In the present study an attempt is made to establish such an experimental technique, by use of a four-element electrostatic probe.

View full abstract

Large eddy simulation of spray combustion was performed with Dynamic SGS model for compressible turbulence. SGS model of combustion reaction was also introduced to take account for contribution of SGS components to combustion reaction rate term. The SGS models were evaluated by comparing results of LES with DNS, which governing equations were not filtered and which had high resolution. A partition method for phase coupling terms and two types methods for partition of gaseous phase variables were applied to compare the results of simulations which have different grid resolutions. The results of LES were greatly improved with the increase of spatial resolution and the contribution of SGS stress decreased with increase of the resolution. However, LES with coarse resolution also could show good agreement with the DNS, in spite of the low resolution. On the other hand, the SGS diffusivity increased with the increase of spatial resolution. From the above results, we could conclude that the proposed partition method for phase coupling terms and partition type a were appropriate for LES of spray combustion.

View full abstract

Two dimensional calculations of the spark ignition process for hydrogen/air or methane/air combustible mixtures were conducted. Two dimensional equations with chemical reactions were solved by fourth-order Runge-Kutta method in Cartesian coordinate system. Harten Yee's upwind TVD scheme was used to capture shock waves produced by a spark discharge. The spark energy was applied into the electrode gap for 1.0 micro second, which is equivalent to capacity spark discharge. Detailed reaction mechanisms for hydrogen and methane combustion chemistries were employed. Equivalence ratio of combustible mixtures were varied, and minimum ignition energy for hydrogen/air or methane/air combustible mixtures was calculated. Early stages of flame development were investigated. Success or failure of the ignition was measured by increase rate in the maxmum temperature after the end of spark discharge. The maximum temperature in the flame kernel decreases after the end of spark discharge and then reaches the minimum value. In the case of ignition, the maximum temperature increases after a certain period of time. In the success of ignitions in hydrogen/air and methane/air mixtures, the maximum temperature in methane/air mixtures is greater that in hydrogen/air ones. Profiles of minimum ignition energy as a function of the equivalence ratio of hydrogen/air and methane/air mixture gases were obtained. Peak values of minimum ignition energy for both of hydrogen/air and methane/air mixtures are observed under lean mixture conditions.

View full abstract

The effects of pressure on outwardly propagating laminar and turbulent flames were studied for propane-air mixtures at the equivalence ratios from 0.8 to 1.3 and the initial pressures from 0.10 to 0.50 MPa. Turbulence intensity, u' was set to 0.80 and 1.59 m/s. The Markstein number of laminar flame was employed to quantify the effects of the equivalence ratio and the pressure of the mixture on turbulent burning velocities. The ratios of the turbulent burning velocities to unstretched laminar ones, u_tn/u_l increased as the mixture pressure increased. Because the unstertched laminar burning velocities, u_l decreased with the mixture pressure, the relative turbulence intensities to the laminar burning velocities, u'/u_l increased with the increase in the mixture pressure. The turbulence Reynolds number also affected u_tn/u_l. u_tn/u_l increased with the increase in the turbulence Reynolds number due to the decrease in the kinematic viscosity with the pressure. And u_tn/u_l increased with the decrease in the Markstein number if u'/u_l, or the turbulence Reynolds number was constant. Turbulent burning velocities were affected by the turbulence Reynolds number, the relative turbulence intensity and the Markstein number relating to the thermo-diffusive effects and flame instabilities.

View full abstract

Dimethyl ether (DME) receives attention as a promising alternative fuel for compression ignition engines and other combustion systems because of its high Cetane Number (almost equivalent to diesel fuel), good combustion-emission characteristics by a high oxygen content, easy handling and productivity etc. In order to make the most of such assets to practical use, it is indispensable to know the laminar burning velocities at wide range of temperatures and pressures and the auto-ignition behaviors. Measurements using a spherical combustion bomb equipped with heaters showed that the pressure and temperature exponents were -0.049, one order of magnitude smaller, and 2.02, slightly smaller than those of methane, respectively, and that the spontaneous ignition temperature is pressure-dependent at equivalence ratio φ=1.0 but turns to pressure-independent at equivalence ratio φ=2.0.

View full abstract

A fluidized bed filter was applied to filtrate particulate matter (PM) in diesel engine exhaust gas. Experiment was carried out using a single-cylinder DI diesel engine to investigate the effects of gas velocity, bed particle diameter and bed height on PM and smoke filtration efficiencies and pressure drops. A theoretical work has been also conducted for further investigation of the effects of the mentioned parameters on PM filtration. The experimental results show that the fluidized-bed DPF provides high PM filtration efficiency and low-pressure drop especially under lower gas flow rates. Smaller diameter bed particles provide higher filtration efficiency under the low gas flow rate ; however, the filtration efficiency drastically decreases with the increase in gas flow rate due to excessive fluidization of the bed particles. The increase in bed height gives higher PM filtration efficiency while it causes an increase in pressure drop. The calculation results indicate that the diffusion collection is a dominant mechanism for PM filtration under the conditions of this study. The lower PM filtration ratios at high exhaust gas flow rates are caused by reduction of the filtration efficiency in diffusion collection.

View full abstract

Coconut oil has high content of straight chain saturated fatty acids such as lauric acid. Also it contains 14.5 wt% of oxygen, which is about 4 wt% more oxygen than that of rapeseed oil. The purpose of this study is to investigate the utilization of coconut oil methyl ester (CME) as a biodiesel fuel for diesel engines. Therefore, the diesel combustion characteristics and exhaust emissions of CME are investigated by using a small DI diesel engine and comparing with rapeseed oil methyl ester (RME) and gas oil. From the experiment results, CME has the better ignitability and the same thermal efficiency as the other test fuels. Both Smoke and NO_x emissions from CME are lower than that from RME and gas oil. Therefore CME can be used as an alternative fuel of petroleum based diesel fuel.

View full abstract