It is known that erosion of fluid machinery can be caused by collapes of a cavitation bubble under high speed flow conditions. To solve this cavitation erosion problem, we performed some experiments on the cavitation process from a subcavitation to a supercavitation stage with a measurement system combining a high-speed video camera and an impulsive pressure sensor. This study focuses in particular on a vortex cavitation bubble in the near wake of a triangular body at the partially cavitating stage which is well known as a highly erosive pattern. Erosion tests were conducted regarding the mechanism of highly impulsive force generation, and bubbl collapsing behaviors were observed. The results show that three characteristic patterns of bubble collapse and erosion occur within the near-wake region.

View full abstract

I present the experimental data obtained using various types of concentric double pipes, in particular, the loss of total head through a turning flow region from the outlet of an annulus between the inner and the outer pipes to the inlet of the inner pipe. They are given in terms of a nondimensional clearance length L/d_i (L is the distance between the annulus outlet and bottom of the outer pipe and d_i is the inner diameter of the inner pipe) for various values of area ratio and Reynolds number. It should be noted that all curves of the cofficient of head loss plotted against L/d_i have the maximum value at about L/d_i=0.625 and the minimum value at about L/d_i=0.250 in the Re range from 4.4×10³∼3.1×10⁴. It was visually clarified that the increase of head loss against L/d_i=0.625 was due to the vortex.

View full abstract

The flow field in rough curved pipes is divided into three regimes, as in rough straight pipes, according to the flow characteristics : (i) Hydraulically smooth regime, (ii) Transition regime and (iii) Completely rough regime. The resistance formulas in the transition regime, which is further divided into two subregimes, and in the completely rough regime are created on the dimensional consideration and with the experimental data. The formulas presented are in agreement with the experimental results, with a reasonable degree of approximation, for various relative roughness and curvature radius ratios of the pipe.

View full abstract

We describe the experimental study of the flow structure past repeated three-dimensional ribs of a square section on a ground plate for various values of spacing ratio S/H. Test ribs of the square section with side length D=23mm and height H=5 mm were aligned at a regular pitch in a square array and in a staggered array, and the spacing between the two adjoining ribs was varied at S/H=2, 3, 5, 7, 10, 13 and ∞ (i.e., a single rib). The time-mean velocity and turbulence intensity were measured using the laser-Doppler velocimeter. As a result, for the case of the square array, there were two flow patterns in the groove. A large recirculation region occupied the entire groove between two adjoining ribs without reattachment for S/H≤5. However, the free streamline leaving the upstream rib enclosed the recirculation region and reattached the ground plate for S/H≥7. On the other hand, a flow pattern independent of S/H was found for the staggered array, which corresponded to the flow pattern for S/H>7 in the square array.

View full abstract

The u and v velocity components of a plane turbulent Momentumless Wake (injection adjust to provide a thrust exactly cancels the model's drag) were measured by the rake of X-wires. The autocorrelations show that large-scale eddies such as karman vortices in a wake are not generated downstream in the Momentumless Wake. The two point double correlations suggest that the momentumless wake is analogous to isotropic turbulence. The fluctuating velocity vectors indicate that the outward extent of the turbulent flows in the momentumless wake is associated with the intermittent blow up and mixing process. Dynamic structures in the Momentumless Wake are investigated by means of the POD. The consideration of the first three modes of the POD allows to precis the instantaneous characteristic structure which plays a dominant role in the dynamical properties of the momentumless wake. The results show that a few dominant wave numbers found in the first eigenvalue of the POD correspond to those of the unstable modes obtained from the linear stability theory.

View full abstract

In this paper the behavior of a turbulent channel flow perturbed from its equilibrium, due to rapid change of pressure gradient, was experimentally investigated. The rapid change of pressure gradient was accomplished by passing an expanded region. Mean velocity profiles, turbulent intensities and Reynolds shear stress following the removal of rapid pressure change were measured by means of a hot-wire anemometer. In the near-wall region, velocity profile and turbulent intensity recover relatively quickly from the change in boundary condition, and the universal law of the wall is maintained except quite close to the expanded region. On the other hand, Reynolds shear stress is significantly perturbed by pressure gradient. Reynolds shear stress decreases in the near-wall region, and increases in the core region.

View full abstract

This paper describes the effect of deformation of a liquid droplet due to its displacement when exposed to air streams. Experiments were conducted using a 60 mm×150 mm cross-sectional and 500 mm long horizontal air-suction wind tunnel. Uniformly sized liquid droplets produced by a longitudinal vibration of the nozzle or by dripping uniform size drops from the nozzle were used. The deformation and the displacement of liquid droplets due to air streams were observed using a stroboscope and were photographed by a camera. The liquid droplets exposed to air streams were deformed ; that is, the longitudinal droplet's diameter which is the normal size to the air stream direction, or the curvature of a drop at the stagnation point, etc are changed. As the projection area of a droplet exposed to air streams is expected to greatly influence the displacement, the drag coefficient, the breakup time, and so on, the longitudinal droplet diameters are precisely measured. The experimental equation for the dimensionless displacement and dimensionless time of liquid droplets submitted in the first report is available for droplets with small deformations. The deformation rate of the longitudinal droplet's diameter as applicable to the experimental equation is investigated quantitatively in detail. The relationship between the Weber number and the maximum/minimum dimensionless longitudinal droplet's diameters is surveyed. The maximum dimensionless longitudinal droplet diameter increases as the Weber number increases. The highest Weber number available to the experimental equation decreases with increasing the droplet diameters.

View full abstract

The behavior of flow in porous media has been investigated using a model of a homogeneous medium obeying Darcy's law. As the permeability reflects the microstructure of porous media, theoretical analysis is performed for evaluation of the permeability of the medium, including the microscopic internal flow. The fluid flow in the medium is assumed to be Stokes flow. Homogenization theory with multiscale asymptotic expansion is used to derive Darcy's law from governing equations for Stokes flow theoretically. The microstructure of the medium is modeled as a body-centered cubic array of sphere particles connected to each other periodically. A cubic unit cell is placed in the microscopic field of the medium to represent a boundary value problem of the internal flow. The internal flow is formulated using intergral equations. Permeability and microscopic fluid flow of the medium are shown by three-dimensional boundary element analysis on the cubic unit cell.

View full abstract

The lattice vortex method (LVM) is used to simulate density stratified flow fields. The point sink flow and the flows past an obstacle (a plate and a semicircular ridge) in a channel of finite depth are calculated. The numerical results agree well with the analytical ones, which indicates that the vortex method is suitable for the simulation of stratified flows. With the LVM, the formation of blocking and lee waves in a stratified flow can be explained clearly and easily. Furthermore, the blocking phenomena and the lee waves are simulated. The blocking area is formed by the development of several eddy regions of cell structures in the stagnant region which are propagated upstream. The lee wave is formed by the columnar disturbances which arise from an isolated source (an obstacle) and are propagated upstream and downstream due to the movement of generated vortices in a stratified flow.

View full abstract

Three dimensional motion of all particles in a centriaugal tumbling granulator was calculated using a discrete element method (DEM). The quality of the products largely depends on the motion of particles in the granulator. Therefore it is important to know the behavior of particles. In the present simulation the velocity profile in a laboratory scale granulator was obtained by applying the periodic boundaries. Furthermore, the effect of the rotor speed on the velocity profile in the particle layer was studied. On the other hand, experiments were made for comparison. In the experiments, particle velocities on the surface of the particle layer were measured by using PTV technique. The particle velocity profile predicted by the present simulation agreed well with the experiments.

View full abstract

A basic study on a newly designed direct-heat-to-power energy conversion device, using a temperature sensitive magnetic fluid, was conducted in order to obtain the fundamental characteristic of the optimized device. The working principle of the device is based on the thermo-magnetic effect of magnetic fluid, in which the driving force (the driving pressure) needed to circulate magnetic fluid in the thermodynamical power cycle can be derived from the magnetization difference in a temperature field. The numerical study used to obtain the basic characteristics of the device is also presented for a cell subjected to the imposed magnetic field and temperature field.

View full abstract

Study on control of vibration of a spring-mass system is conducted. A cylinder forced to vibrate is filled with a magnetic fluid and a piston immersed in the magnetic fluid is attached to the mass. The magnetic field necessary for stopping vibration of the mass is predicted. Validity of the prediction is proved by numerical experiments, where equations of motion for the spring-mass system and the axisymmetric flow of the magnetic fluid are simultaneously solved. The magnetic field generated by a couple of coils can effectively damp transient vibration and control steady vibration to be extremely small.

View full abstract

The transportation mechanism of fluids and/or solid films proposed in the previous papers is achieved by coupled waves which propagate along a thin elastic membrane covering a magnetic fluid layer in a shallow and long rectangular vessel. In order to control the propulsion system and to improve its efficiency, it is necessary to adjust the frequency and wavelength of waves by considering the dispersion relation of the coupled system consisting of a magnetic fluid layer and an elastic membrane. In this paper, transient coupled waves generated by applying external impulsive magnetic forces on the concentrated small region of the magnetic fluid layer are investigated experimentally and partly analytically. Various dispersive waves observed depend on the geometrical and physical parameters such as density of the magnetic fluid, depth of the fluid layer, and kind and thickness of the membrane.

View full abstract

Theoretical and experimental investigations were performed on the structure of a correctly expanded supersonic laminar jet from axisymmetric nozzles. A small perturbation method was used in the theoretical calculation, and the effect of the Reynolds number Re_De at the nozzle exit on the Pitot impact pressure on the jet center line is discussed. A low-density wind tunnel was used for the experiments with the Mach number M_e at the nozzle exit of about 2.9 and Re_De from 1000 to 1800. The axial and radial Pitot impact pressure profiles were measured, and it was found that the Pitot impact pressure distributions on the jet center line correlate quite well with the parameter x/(D_eRe_De), where x is the downstream distance from the nozzle exit and D_e is the nozzle exit diameter. The present theory suggests that the length of the jet potential core increases linearly with the increase in Re_De, and this result is found to agree quantitatively well with the results of the experiment.

View full abstract

The three dimensional complex turbulent flow fields induced by a four flat blade paddle impeller in agitated vessel were measured by laser doppler velocimetry. Mixing vessel used was a closed cylindrical tank of 490 mm diameter with a flat bottom and four vertical buffles, giving water volumes of about 100 1. The impellers were at the midheight of the water level in the tank. A height of liquid (water) was equal to the vessel diameter. Three components of mean velocity were measured at three vertical sections θ=7.5°, 45°and 85°, in several horizontal planes. Mixing Reynolds number N_Re was 1.2×10⁵. It can be found from the results that circumferential mean velocity profiles show the symmetrical shape in the upper and lower sides of impeller. Secondary velocity components, such as axial and radial velocities, however, were not in symmetry. For this reason, the ratio of circulation flow volume which enter in upper and lower sides of impeller was roughly 7/3. In both the middle and buffle regions, mean flow velocities (flow patterns) were different, dependent of three vertical planes with different circumferential angle measured from buffle.

View full abstract

Sport science progresses step by step in the world. This work is a challenge to develop the athlete woman swimming suit with low fluid drag. To begin with, the fluid drag of the woman swimming suit is very small. It is very difficult to measure the several percent difference in the fluid drag of the swimming suit. Special experimental apparatus is developed to measure the fluid drag, precisely. It can successfully measure the fluid drag of athlete woman swimming suits at the precision 1∼2%. As a result, the cloth with low fluid drag is found. It is worked water repellent into every other stripe on the cloth. The cloth is woven of thin threads (polyester 80% and polyurethane 20%). Also, the relationship between fluid drag for the model body and the water depth from the water surface to the model body is investigated in details.

View full abstract

A full three-dimensional simulations by means of a finite difference method of incompressible near-stall flows in a stationary straight cascade of blades is presented in this report. The aim of the work is to compare the flows with those in a rotating cascade which is described in the accompanying paper. Two cascades are nearly identical in geometry with respect to inlet flow condition, namely, same attack angle which is uniform in spanwise direction and also to flow in tip clearance. It is intended in these works to examine the validity of two-dimensional simulation of stalled flow in which flow-normal vortices play key role. The basis of the presumed flow pattern is too much different from measured flows and it is required to clarify the mechanism of the difference, in order to make two-dimensional simulation valid. The simulation presented here suggests that the flows, even if they are three dimensional. are basically similar to two-dimensional cascade flows but are quite different from those in the corresponding rotating cascade, due to centrifugal force applied in the latter.

View full abstract

Full three-dimensional simulations of incompressible stalled and stall-free flows in a single stage axial rotor with and without tip clearance are presented in this report. The simulation are conducted for one flow passage between blades assuming periodicity in circumferential direction, and hence rotating stall is not considered presently. Effect of centrifugal force which manifests itself seriously in the flows in axial rotors, particularly in stalled flow is of present concern. Comparing with the counterpart flow in stationary straight cascades for similar inlet flow conditions which are mentioned in the accompanying paper, basis of models required to make two-dimensional simulation which is far small in computational load possible to predict onset of unstable fluctuating flow in an axial rotor are expected to be established. It is revealed that due to centrifugal force, separated flow at the blade leading edge is pushed toward casing wall and sheds intermittent vortex tubes, inducing high fluctuation in the flow. Unsteady separation with fairly long period is crucial in the understanding of the flow in the case of stalled rotor.

View full abstract

A guantitative investigation is carried out on the generation of a compression wave in highspeed train tunnels and on its transition to a shock wave. Discussions are based on results of experiments which were conducted in a scaled train tunnel simulator. The overpressure level of the compression wave is formulated using a modified Hara-Maeda formula, in which an effective blockage factor is taken into account. The length of the shock wave transition region is estimated from the pressure gradient of the compression wave at the tunnel entrance. The installation of a hood at the entrance is found to disperse the pressure rise, thereby the length of the shock wave transition region is increased. However, the length decreases significantly with an increase in the train speed. Conditions under which the entrance hood works effectively are formulated. Consequently, it is concluded that a critical train speed exists beyond which the entrance hood becomes no longer effective for retarding the transition.

View full abstract

In this work, the airflow generated by several types of winds in a subway station is numerically simulated by the two-dimensional SIMPLE method. A steady wind, an unsteady wind and a wind created by movement of actual train, which are the inflow or outflow conditions in the tunnel of the station, are treated for evaluation of the steady or unsteady effects of airflow in the station and estimation of the accuracy of this analysis. The subway station model used in this analysis consists of a simplified configuration of a platform, a baffle at the front end of the platform and four tunnels. In the case of the steady inflow, the airflow circulates around the platform behind the baffle. In the case of the unsteady inflow, this circulating flow gradually extends with time from the front end of the platform to the center. However, this circulating flow does not appear in either case of the steady and the unsteady outflow. In the case of the actual train wind, the calculated velocity of time process agree well with the results of fieldwork tests in the tunnels and on the platform. This numerical analysis, therefore, can be used to evaluate the airflow in an actual subway station with an intricate configuration including for example staircases and a ventilation system.

View full abstract

The turbine bias angle, one of the major design parameters of a three-element automotive torque converter, is investigated in terms of its influence on internal flow and performance. The internal flow through the three elements of a torque converter at a speed ratio of 0.8 is computed using a three-dimensional CFD code. The flow field within the turbine blade passage is analyzed in detail for three turbines with different bias angles. The results show that a large friction loss or deceleration loss is generated in a turbine with an excessively small or large bias angle, leading to performance deterioration. A correlation is also seen between the overall torque converter efficiency and the turbine efficiency. The computed overall efficiency and the secondary flow pattern near walls of the turbine passage agree well with the tendencies of the experimental results obtained using actual machines in terms of the bias angle.

View full abstract

This is the second report of study on dynamic behaviors of a pump system by using Bondgraphs method. In this paper, the more detailed model of the system bondgraphs on the dynamic behaviors has been established considering the influence of an impeller configuration and internal losses of the pump. Especially, the mathematical expressions of nonlinear internal losses of the pump are derived and their effects have been investigated. It is known from this research that in high speed transients the impeller configuration affects the dynamic behaviors greatly. And the internal losses become important in low speed transients. Also, through this study, the effectiveness of Bondgraphs method to the transients analysis for turbomachinery system is proved.

View full abstract

A series of studies on dynamic characteristics of noncavitating centrifugal pump were extended to cavitating case. Experimental study was carried out on transient behavior of cavitating centrifugal pump at the sudden opening/closure of discharge valve. Cavitation behavior was visualized during transient period of centrifugal pump by using high speed video camera. And instantaneous pressure and flow rate were measured at suction and discharge ports with rotational speed during the transient period. Unsteady pressure as well as flow rate was related to the time-dependent cavitation behavior. As a result of the present study, pressure fluctuation was found to occur due to water column separation or oscillating cavitation at rapid transient operations.

View full abstract

In the 1st report, the transient characteristics of a cavitating centrifugal pump were related to the cavitation behavior at sudden opening or closure of discharge valve. In the 2nd report, experimental study was extended to transient behavior of cavitating centrifugal pump at the quick starting/stopping of pump. Unsteady pressure as well as flow rate was related to the time-dependent cavitation behavior in the manner similar to quick discharge valve operation case. As a result of the present study, pressure fluctuation was found to occur due to water column separation or oscillating cavitation at rapid transient operations. And experimental results indicate that transient phenomena at opening of discharge valve and pump starting are quite different from those at closing of discharge valve and pump stopping.

View full abstract

Experimental data in present authors' first and second reports on transient characteristics of cavitating centrifugal pump were rearranged to get a criteria for the occurrence of the fluctuations of pressure and flowrate at a variety of transient operations. Transient behaviors were classified to the increasing and decreasing suction flowrate cases at transient operations. One dimensional unsteady flow analysis was applied to the transient phenomena and the calculated critical conditions for the occurrence of the fluctuations agree qualitatively with measured ones. The criteria for the occurrence of the fluctuating pressure and flowrate was presented for the transient phenomena during sudden opening/closure of discharge valve as well as starting/stopping of pump, by using newly introduced unsteady parameter.

View full abstract

Flow and temperature fields in falling liquid films with interfacial waves have been obtained by means of numerical simulation in which the algorithm is based on the HSMAC method and interfacial boundary conditions are treated with newly proposed methods. Small amplitude disturbance waves at a low frequency develop into isolated solitary waves which are composed of a roll wave and capillary waves. Waves disturbed at a high frequency interfere with each other and develop into disordered waves. Circulation is observed in the roll wave, while there is no circulation in the disordered waves. Temperature fields in the wavy film are distored by the convection effects and differ greatly from those in the laminar film. The circulation in the roll wave has an especially strong effect on the temperature fields. The interfacial waves enhance the heat transfer by two kinds of effects : the variation of the film thickness and the convection in the film. The dominating effect depends on the Prandtl number.

View full abstract

The present paper describes experimental results on the fluid flow and heat transfer of natural convection in between the upward-facing, horizontal heated plate and the insulated cover plate. The experiments were carried out with water. The widths of the test plates W and their gaps H were changed as W=50∼250 mm and H=10 mm∼∞ (no cover plate). The visualization studies with dye and liquid crystal thermometry revealed that the roll cells of which axes parpendicular to the flow direction appear and become dominant over the heated plate with decreasing the gaps. These roll cells inhibit the heat transfer and, thus, the heat transfer coefficients become smaller than those without cover plate. While, it is found that the flow and heat transfer in the region near the plate edges are unaffected by the cover plate. The conditions of the above reduced heat transfer are determined empirically. Moreover, the non-dimensional correlations for the local and the over all heat transfer coefficients of the heated plates are proposed based on the present heat transfer results.

View full abstract

The experimental study of body forced convection of pure vapor on a vertical surface (l=2980 mm) is presented. CFC11, CFC113 and HCFC123 are used as the working fluids. The flow patterns of condensate film are classified into four kinds of flow : L flow (laminar flow), L-S flow (laminar and sine wave flow), L-S-H folw (laminar, sine wave and harmonic wave flow), L-S-H-T flow (laminar, sine wave, harmonic wave and turbulent flow). The nondimensional correlations for predicting the average heat-transfer coefficients are obtained for four kinds of flow and compared with the experimental results, respectively. Assuming more simple four kinds of flow patterns, which are L flow, S flow (sine wave flow), H folw (harmonic wave flow) and T flow (turbulent flow), simple nondimensional correlations for predicting the average heat-transfer coefficients are also obtained for S flow, H flow and T flow. The nondimensional correlations for combined H and T flow are also obtained.

View full abstract

This paper deals with heat energy release characteristics of shape-stabilized latent heat energy storage particles packed into a fluidized bed-type cylindrical heat energy storage vessel. This type of latent heat energy storage material consists of normal paraffin (pentacosane C₂₅H₅₂, latent heat : 164 kJ/kg, melting point of 327.2K) as a latent heat material, and polyethylene as a shape-stabilizing material. The effects of cold inlet air temperature and flow rate, and the amount of heat energy storage particles on the outlet air temperature from the heat storage vessel and the completion time of the heat energy release process were investigated. As a result, the non-dimensional correlations for the outlet air temperature from the heat storage vessel and the completion time of the heat energy release process were expressed in terms of non-dimensional pumping power, Stefan number and ratio of the packed particle layer height to the diameter of the cylindrical heat storage vessel.

View full abstract

The fin efficiency of serrated fins was analyzed and an analytical solution was derived as a function of modified Bessel functions. Two assumptions, i.e., thermal insulation at the end surfaces of segmented sections, and a uniform heat transfer coefficient over the fin surface, were employed in this analysis. To determine the effect of these assumptions, a heat transfer experiment was performed. From a comparison of the experimentally evaluated fin efficiencies with the analytical solution, a correction factor was obtained for a typical serrated fin configuration.

View full abstract

This study deals with gas sensible heat-radiative energy exchange in fibrous media used for a system of energy recovery from high temperature exhaust gas from an industrial plant. The effects of fiber diameter, orientation, medium thickness and the fiber radiative characteristics on the recovered energy have been numerically investigated. The results show that, if the heat transfer coefficients between the fibers and the gas are assumed to be those of circular cylinders in a cross-flow, the gas sensible heat can be exchanged to radiative energy and recovered by use of a medium with fibers less than 500μm in diameter under the present calculation conditions. The results also show the value of the absorption index of the fiber material required for efficient energy recovery.

View full abstract

A method for predicting the natural convection air cooling performance of notebook size electronic equipment was investigated experimentally and analytically. Experiments were performed on the natural cooling performance for all of the heat dissipation paths in a typical notebook size electronic device. These paths include the surfaces, which are modeled as a flat plate with local and uniform heat sources and a keyboard with a uniform heat source, both open to the atmosphere, and the components inside the housing, which are modeled as a locally heated flat plate in an enclosed space. Total heat dissipation from a flat plate can be calculated from general formulas for radiative and natural convective heat transfer. New formulas were proposed for the convection heat transfer of the keyboard and the locally heated flat plate. An analytical model for predicting the temperature distribution in a notebook-sized device is developed as a combination of thermal-resistance network analysis and the new correlation formulas. As a test, the temperature distribution of an actual personal computer was predicted using this model within an accuracy of 15%.

View full abstract

This paper proposes an innovative coal combustion/gasification furnace incorporating pebble bed filter for high temperature molten slag removal. The experimental demonstration of this high temperature filter is reported. Molten coal slag paricles generated by injecting fly ash of coal into high temperature natural gas fired combustion gas are successfully trapped and extracted from the filter, and 97.7% slag removal efficiency is demonstrated. Alumina pebbles receive significant attack from the reaction with the molten slag under the temperature around 1700°C, while this attack can be greatly suppressed by lowering the operating temperature of the filter below about 1600°C. A new coal gasification combined cycle power generation system employing the pebble bed coal gasifier is proposed, whose thermal efficiency is expected to reach as high as about 50%.

View full abstract

Molten carbonate fuel cells (MCFCs) are expected to be large-scale, highly efficient energy sources in the future. A reactant gas seal is one of the critical issues in achievement of reliability of the MCFC stack, which is a key piece of equipment in an MCFC power plant. At present, the wet gas seal with carbonate molten at the operating temperature (650°C) is generally considered to be suitable as such a seal. On the other hand, to be cost-competitive with existing power plants, the MCFC stack must consist almost entirely of thin sheet metal parts. In this report, a wet gas seal configuration composed of thin sheet metal parts is proposed. Then, the long-term performance of the wet gas seal in an actual MCFC is evaluated.

View full abstract

It is useful to analyze the internal temperature distribution of a molten carbonate fuel cell (MCFC) stack for the purpose of realizing the high performance and stable operation. In this paper, we developed a network model that can analyze pressure, current density and temperature distribution for cathode gas counter-flow type stack. The calculation result showed good agreement to the experimental temperature distributions, so this analysis model could be an effective method to analyze the stack performance. Using this model we proposed the gas flow type stack that can realize the uniform temperature distribution. The proposed gas flow type stack showed higher cell voltage because the higher average temperature has been accomplished.

View full abstract

The extinction limits and flammable regions of a nonadiabatic counterflow premixed flame at a large Lewis number are investigated numerically. It is found that interaction between the radiative heat loss and the Lewis number effect results in four kinds of flame bifurcations. Furthermore, the results show that there exist four different kinds of flame regimes : normal flames, radiation-stretch-induced weak flames, radiation-Lewis number effect-induced weak flames and unified weak flames at typical equivalence ratios. An increase in the radiation intensity reduces the extinction limit of a normal flame but extends the flammable region of weak flames to large stretch rates. A general graph showing the extinction limits and the flammable regions is obtained. It is found that the G-shaped curve obtained at a low Lewis number is reduced to a K-shaped curve at a large Lewis number. This G-K transition presents a good explanation for the experimental results obtained in microgravity experiments.

View full abstract

Lewis number effect on the local quenching phenomena of lean methane/air and propane/air turbulent premixed flames formed in a wall stagnation point flow have been studied experimentally. The stretch rate due to flow divergence was 60 s-1, while the turbulent intensity of velocity fluctuation in the approach flow was 0.38 m/s. The flame front behavior was taken by using laser tomographic technique with a 16 mm high speed camera. Near the extinction limit, the flame front reaches the cooled wall. Then, the flame is locally quenched, though it is globally stable. The local quenching occurs when the concave curvature to the unburnt gas is close to the cooled wall. The local quenching is sensitive to the lewis number in that it's probability of lean methane/air flame (Le<1) is more than ten times that of propane/air flame (Le>1). This suggests that the interaction of the cooled wall and flame front plays an important role in the local quenching of lean methane/air flame. When the cusp flame front reaches the cooled wall, the duration of local quenching time is short. In this case, the local quenching area may be small, and the global extinction does not occur. On the other hand, when the small curvature flame front reaches the cooled wall, the time is long. Then, the local quenching area may be large, and the local quenching become a trigger to develop the global extinction.

View full abstract

A circulating flow combustion device was developed for a spontaneous ignition experiment on fuel gas jets under high pressure. The combustion device is composed of a circular duct, a fan and electric heaters. Ethylene was injected perpendicular to a hot air stream of high pressure circulating through the duct. The ignition delay was measured and the effects of pressure, air temperature and fuel jet velocity were examined. The results show that the ignition delay time τ is expressed in terms of pressure P_a and air temperature T_a by the equation τ=APⁿ_a exp (E/RT_a), and the apparent activation energy E is constant, unrelated to pressure. In this study, E was 243 kJ/mol and the pressure exponent n was -1.3.

View full abstract

A flash photolysis technique is useful to study chemical reactions of explosively burning hydrocarbon/oxidizer mixtures associated with emission and absorption spectra observations. When a single ion probe is inserted into the reacting gases to measure the ion current. it is positively or negatively electrified for a moment according to the conditions at the beginning of chemical reactions. A vacuum ultra-violet light (VUV) appears together with the CH, OH and C₂ spectra and causes a photoelectric effect not only on the probe surface but also on the inner surface of the quartz combustion tube. The positive or negative electrification can be derived theoretically from the photoelectric effect on the quartz tube. It is concluded that the emitter is highly excited carbon monoxide, because the VUV photon should have an energy higher than 10 eV to cause electron emission from the quartz tube inner surface.

View full abstract

A portion of the heat accumulation-type radiant-tube burner, which has a pair of gas burners with a honeycomb regenerator on both ends, was utilized as a test rig for spray ignition. The effects of various parameters on the ignition process were examined by perpendicularly injecting a liquid fuel (gas oil) into the hot-air stream immediately after cutting off the city-gas supply to the burners. If a cavity is embedded in the wall confronting the injection nozzle, only the jet portion of the spray is exposed to the hot stream and ignited, while the tip portion chills in the cavity. With no cavity, on the other hand, ignition takes place in the tip portion just before impinging on the wall, or the secondary spray generated by the impingement is ignited within the wall boundary layer. The data for both cases and our former shock-tube data were compared with those of other researchers and examined. As a result, the reason why ignition delay data were so widely different from researcher to researcher was elucidated to some extent.

View full abstract