The steady motion of a small particle passing through a sinusoidally curved channel with a constant width is numerically studied. The equations governing the fluid and the particle formulated in the curvilinear coordinates are solved by using the finite difference method. The fluid-particle interaction is assumed to be described by the Stokes drag law. The trajectory of a small sphere with a given radius and weight steadily passing through the channel is found to be only one type by means of rewriting the inertia term in the equation of particle motion in the form of the spatial derivative. The numerical solution of the equation of motion in this form has much less computational error compared with that based on the traditional method treating this problem as an initial value problem. The effects of the four parameters, the Reynolds number, the aspect ratio, the Stokes number and the wave number of the wavy channel on the steady trajectory of the particle are treated in detail. It is found that the solution obtained numerically coincides with that had obtained analytically by the authors. The difference in the motion between the particle and the fluid is clarified by comparing the trajectories of the particle and the stream line with the value of the half flow rate.

View full abstract

The motion of small particles in turbulent mixing layer was experimentally studied in order to clarify the dominant factors on particle motion in turbulent gas flow. Spherical glass particles were loaded into the initial point of a two-dimensional air mixing layer. The laser-Doppler anemometry with particle size discrimination enabled to perform measurements of both particles and gas-phase velocities and also particle number density. The results show that the particle dispersion depends strongly on the stokes number, that is, the ratio of the particle relaxation time to the characteristic time scale of the large scale eddies in the mixing layer and overshoot phenomena were observed in the range of the stokes number of about 1.

View full abstract

Experiments were performed in a two-phase, two-freestream plane mixing layer to study the behaviour of heavy spheres in an intermittent shear flow. Particles with three different diameters were employed in these experiments. Comparisons are presented between data on streamwise and transverse components of particle and gas velocities. These data, as well as flow visualization images, suggest that the ratio of the particle time-constant τ_p and the large-scale eddy turnover time plays an important role in the cross-stream transport of a heavy dispersed phase. When the τ_p/τ was approximately 0 (1), heavy particles began to move with gas flow. There exists movement of heavy spheres back toward the particle feedstream. The transverse component of large particle turbulent velocity, v', in the interior of the layer was observed to be significantly smaller than the corresponding streamwise component, u'. <v²>/<u²> reaches an asymptotic value, 0.2, around the high-speed side of the mixing layer.

View full abstract

Present paper treats the cavity flows of elastic symmetrical wedge by singularity method and clarifies the notable inter-relationships of cavity flow models to beam stiffness on cavity characteristics. The main results are summarized as follows; (1) While using the analytic relations for the cavity length and cavitation coefficient, the beam deflection is determined by iteration procedures from bending equation and then the cavity characteristics such as the pressure difference between the in-and out-sides of wedge, the cavity length and also the cavity drag are finally obtained. (2) As an typical example for exter-and inter-nal flows, the cavity flows in gravity field and passage between parallel walls are examined respectively: Some notable relationships of beam stiffness to cavity flow models, in particular the reduction of cavity drag by beam deflection are pointed out concreately through some numerical examples for each cases.

View full abstract

The two-dimensional flow of an incompressible viscous fluid past a circular cylinder placed in a uniform shear field is analytically solved with different Oseen's velocity fields considered to avoid the "Garstang paradox". The general solutions of the corresponding velocity consist of the nonsymmetric solution to the Oseen equations and one to the equations of the perfect fluids for shear flows past a circular cylinder. The constants of the above are determined by the viscous condition on a circular cylinder. On the basis of these solutions, the expansion formulae of the stream function, Ψ, and the vorticity, ζ, are obtained up to the fourth power of the Reynolds number, Re. The flow patterns for values of a shear parameter, ε, in the range 0≤ε≤0.5 are depicted by utilizing these expansion formulae. They are shown with good results at low Re. Also, the expansion formulae of the lift, moment, and pressure distribution are obtained up to the fourth power of Re. Lastly, successive approximations for the ε-term based on the above solutions are carried out and discussed in the same manner as in the previous paper.

View full abstract

Experiments were carried out using a circular model with a wake splitter plate having length ratio (L/d)=1/6, 1 or 3 in the Reynolds number range of 5×10⁴≤Re≤10⁷ using a cryogenic wind tunnel. The resulting variations of pressure distributions, drag coefficients, back pressure coefficients, the minimum values of pressure coefficients and Strouhal numbers with the length of wake splitter plate are presented. The effect of the wake splitter plate on the drag and Strouhal number are shown to change depending on the Reynolds number range and on the length of the wake splitter plate. Effects of the wake splitter plate on the reduction of the drag coefficient is most remarkable in the subcritical Reynolds number range. In the supercritical Reynolds number range the wake splitter plate serves to suppress the separation bubble. In the transcritical Reynolds number range the effects of the wake splitter plate can hardly be recognized.

View full abstract

The discharge coefficient of a circular orifice set concentrically to the bottom of a circular cylinder vessel was studied experimentally in the flow into the air under the head lowering with the time. The discharge coefficient of orifices C_d was determined by the conventional formula and the effect of several factors (Reynolds number Re, Weber number W_r, diameter ratio D/d, thickness-diameter ratio h/d and dimensionless head H/d) to C_d was discussed. It was found that C_d decreased with an increase of Re under a constant value of W_r, C_d was almost independent of D/d and C_d increased with an increase of h/d. Besides, we tried to formulate the effect of some factors (Re, W_r and h/d) to C_d and proposed an experimental formula in the range of Re=1100-14500, W_r=100-700 and h/d=0.1-0.4.

View full abstract

The third-order upwind finite difference scheme is used for the convective terms in three-dimensional Navier-Stokes equations to simulate directly the free jet flow issuing from a slit nozzle. The calculated flow field ranges from the nozzle exit to the fully developed turbulent region. The kinetic energy conservation scheme is also devised for the convective terms. The calculated results of mean velocity distributions and velocity vector diagrams are compared with the experimental data. The mesh width is roughly ten times Kolmogorov microscale; however, the mean velocity distributions in the free jet are in fairly good agreement with the experimental data. This is because the simulation well represents the motion of large-scale eddies which play an important role in a flow field formation. Hence the neglection of the small-scale eddies in the third-order upwind finite difference scheme does not largely affect the simulation of large-scale eddies.

View full abstract

An experimental study was made to clarify the fundamental characteristics of the tube side flow rate distributions in a multi-tube heat exchanger. Experimental parameters considered were the flow rate of feed water, the nozzle direction, the nozzle diameter, the length of the branch tube and the length of the header. The velocity in each branch tube was determined by the frictional pressure drop measured. The results show that both the dynamic pressure of feed water to the heat exchanger header and the pressure loss in a branch tube had a significant effect on the flow rate distribution. In some experimental conditions, the maximum flow rate in a tube was approximately twice as large as the mean flow rate of the tube bundle, which may not be ignored in the design of heat exchanger.

View full abstract

A discrete vortex model is introduced in the calculation of the starting flows past a two-dimensional elliptic airfoil oscillating in pitch at large incidences. The potential flow past an obstacle is represented by a fixed number of bound vortices distributed along the surface and the viscous wake by an increasing number of free vortices, which are shed from the leading and trailing edges with a regular time interval and convected downstream under the action of the free stream velocity and the vortex-induced velocity. Every time stage of the calculated wake is visualized by means of point vortex plots and streamline tracings. In parallel, the instantaneous hydrodynamic loads exerted on the oscillating airfoil are estimated by using the Blasius formulae for the unsteady regime. The tested experimental parameters are the reduced pitching frequency ranging between 0.1 and 1.0 and the amplitude of oscillation between 7°and 15°. The mean incidence is fixed at 30°and the initial incidence at the minimals. Some comparison is made with the purely experimental streamlines presented in the first report.

View full abstract

The flow pattern in the oil separator of a rotary vane compressor is studied by the finite difference method. Three-dimensional turbulence flow is calculated in the generalized coordinate. Pressure for the flow field is calculated by the MAC method, and time integration is solved by the Euler's backward scheme. The third-order upwind finite differencing scheme is applied for the convective term in the Navier-Stokes equation. All computations are performed on a FACOM VP 200 supercomputer system. The results of this investigation show that the numerical design is available.

View full abstract

A Simple type (Model 2) of the turbulent stress model developed by Launder, Reece and Rodi is adopted changing to a high-Reynolds-number turbulence model, to estimate the accelerated effects for double conditions: convex curvature, and adverse pressure gradient. To evaluate the accuracies of the model and the numerical computation method, the velocity profiles, turbulent stress profiles and quantities of turbulent energy production terms are calculated, and compared with the experimental values for CASE IIP. The condition CASE IIP consists of two conditions: convex curvature radius R=500mm, and weak adverse pressure gradient dc_p/dx=0.2674(1/m). As a result, it is clarified that the effects of the double conditions are more accelerated than the total of the individual effects, and the boundary layers on the convex surface are highly developed.

View full abstract

A new higher order method is devised for the numerical simulation of two-dimensional homogeneous isotropic turbulence. The spatial derivatives of the Navier-Stokes equations are discretized by means of the modified differential quadrature (MDQ) method. The resulting system of ordinary differential equations in time is integrated by the fourth-order Runge-Kutta-Gill (RKG) schemes. The elliptic (Poisson) equation in solved by a new variable-order multigrid method. The direct simulations are presented at resolutions up to 1024×1024. The results suggest that the present higher order method is more efficient, and that the history of the vorticity field consists of three stages, i. e., development, peak, and decay. The corresponding power laws of inertial energy spectrum are k^-4, k^-3, and k^-4, respectively.

View full abstract

This paper describes the formation and decay of vortex shedding, turbulent eddies, and turbulent wakes from a square cylinder of finite length placed on a ground plane. The experiment was carried out in an N. P. L. blowdown-type wind tunnel having a working section of 500mm×500mm×2000mm in size at the Reynolds number 2.5×10⁴. As a result, it was found that (i) the vortices shed from a square cylinder are arch vortices for H/D=1 and 2, and change into Karman vortices at H/D=4∼7; (ii) the number of turbulent eddies of a large scale contained in the turbulent near wake increases as the downstream distance increases, and the scale of eddies in the near wake enlarges with an increase in the value of H/D; and (iii) the effect of the free end of a square cylinder on the near and far wake was determined from the velocity measurement.

View full abstract

The velocity profile and the distribution of the turbulent stresses in a curved tube drastically differ from the simple profiles in a straight tube due to the secondary flow induced by the centrifugal force. Characteristics of this complex flow are numerically analyzed by using the modified k-ε turbulent model equation combined with the algebraic stress model. Two important Reynolds stresses can be determined from the algebraic turbulent stress equations. They are closely related to the turbulent generation. The proposed model includes a new coefficient which should become zero in the ordinary k-ε model. Its value is determined so that numerical results might agree with the experimental results.

View full abstract

An experimental study is made for the turbulent boundary layer on the rotating cone in the axial uniform stream. Measurements of the mean velocity distributions were made at various values of the surface-to-free stream velocity ratio λ_b for two different cone vertex angles, 0=15°and 30°. In the polar diagram of the mean velocity distributions, the profiles are not only dependent on the local velocity ratio λ but also on θ and R_w, where R_w is the Reynolds number based on the wall velocity. The distributions of the swirl flow Richardson number reveal that the destabilizing effect of the centrifugal force is considerably stronger at θ=15°than at θ=30°. Discussions of the logarithmic velocity distributions are made for the azimuthal and the meridian velocity components as well as for the resultant velocity.

View full abstract

A gas-liquid two-phase flow in a capillary tube or a narrow passage is encountered in many applications, for example, the flow of coolant in evaporators of many kinds of heat exchangers. Sufficlent information on fundamental parameters, such as void fraction and pressure drop, however, are not available in designing such devices. In the present investigation, we made a detailed observation of the flow pattern of an air-water two-phase flow in a horizontal capillary tube with an inner diameter of 1.0mm to 4.9mm, and attempted to correlate and explain the behavior of the pressure drop changing with the flow rates of both phases by introducing a physical model.

View full abstract

Void drift is one of the main flow interactions between subchannels. It results from the tendency of a two-phase flow to approach an equilibrium condition without a transverse pressure differential. In this paper, void drift is phenomenologically described. Experiments were performed for two-phase flow of air and water mixtures through a multiple channel consisting of two interconnected subchannels. Data on the developing flow redistributions of both phases are presented. The "void settling model" proposed by Lahey et al. is also studied by use of the present data.

View full abstract

A method combining characteristics and finite difference was developed by Zielke to predict a transient phenomena of a fluid transmission line. This method is easy to apply to complicated systems and has high accuracy and is, therefore, used frequently. However, it requires enormous computation time and huge computer storage to simulate frequency-dependent friction in transient liquid flow. In order to counteract these disadvantages, the authors paid attention to the fact that the weighting function which causes the above problems, is composed of two parts, i.e. the part expressed by exponential functions and the part expressed by other functions. In order to perform the mathematically identical calculation without approximations, a new method has been developed which requires much less computation time and computer storage than the conventional method. The operation has become easy by showing the calculation process as a block diagram for a digital filter because this calculation is the same as the calculation of a digital filter.

View full abstract

A method of flow channel design of multi-cavity molds for thermosets has been developed. This program computes runner and gate dimensions such that the pressure drop during filling from a pot to each cavity is the same in order to obtain equal flow rate in each cavity. Viscosity changes were estimated by using a method for analyzing mold filling dynamics in circular channels. Flow channels of an actual mold for encapsulation of semiconductor devices were designed by using this program. As a result, well balanced filling was achieved. The behavior of balanced filling was found to yield a great reduction in defects such as wire sweep and void formation in molded devices compared with those produced by a conventional mold.

View full abstract

Predictions of pool void fraction in bundle passages are important in the estimation of dryout elevations in a core or a steam generator during a loss-of-coolant accident (LOCA) of a nuclear reactor. Many studies on the void fraction in circular pipes and rectangular channels have been performed: however data on pool void fractions in the bundle passages is scarce. The void fractions in four bundles of different geometries and a rectangular channel are measured by using an X-ray CT scanner in the present air/water experiment. The effect of bundle geometries on the pool void fraction is studied compared with data on the rectangular channel without bundles.

View full abstract

As a method of concentration and velocity measurements in gas flow, hot-wire techniques are proposed. The device is constructed of an ordinary hot-wire and a concentration probe which consists of a sonic nozzle and a hot-wire probe. The probes should be 2mm apart from each other to avoid the mutual influence. It may be possible to detect concentration and velocity simultaneously in the turbulent mixing region of two different gases, without introducing particles into the fluid with this method.

View full abstract

A method of simultaneously measuring concentration and velocity in the mixing region of two different gases was illustrated in the 1st report. Calibration for each probe becomes important for the analysis. The actual measurements are carried out in the turbulent region formed by axisymmetric jets issuing six kinds of different gases of varying densities. The results of the experiment give some new conclusions concerning the characteristics of the free jet. The present paper deals with the axial decays of concentration and velocity through examination of the mean and fluctuating values.

View full abstract

The stable finite-element techniques for the unsteady high-Reynolds-number flows are discussed by solving a rotating cone problem of the advective equation. High-order time-marching methods are used, e.g., the second-order or the third-order Adams-Bashforth method, the predictor-corrector method, the four-stage Runge-Kutta method and the balancing tensor diffusivity (BTD), which is derived from the second-order explicit Euler time integration. The accuracy of the multi-pass explicit algorithm and simple one-point quadrature are also discussed.

View full abstract

A new procedure for a numerical analysis of incompressible potential flow through cascades of airfoils with arbitrary cross-sections has been presented. The procedure consists of three steps. First, airfoil rows in a physical plane are mapped into near-circle rows using a simple mapping function; then the flow field through the near-circles on the mapped plane is calculated with the singularity method. Finally, the flow on the mapped plane is transformed to that through the cascades of airfoils. The method for determination of a parameter of the mapping function to obtain near-circles with sufficiently smooth contours on the mapped plane, which is indispensable to achieving the highest accuracy in the calculation with the singularity method, has been proposed. The validity of this procedure is demonstrated by comparing the present results with exact solutions, which indicates that the method proposed is very useful for the numerical analysis of the cascades flow with simple thin airfoils as well as many practical airfoils with a camber and thickness.

View full abstract

The performance of the compressor-type cascade blades of small aspect ratio (AR, AR=2.08, 1.39, 1.04 and 0.83) at the center of the blade span has been experimentally assessed. The relation between the performance and Reynolds number (Re) is obtained. This relation is used to eliminate the Re effect on performance. Consequently, pure effects of blade AR on performance, i. e., total pressure loss, turning angle and axial velocity ratio, are obtained.

View full abstract

Unsteady aerodynamic forces on a turbine cascade blade due to a sinusoidal gust have been studied experimentally. Turbine blades having large camber, large thickness and large turning angle, it is difficult to obtain unsteady forces theoretically. In spite of many theoretical investigations having been carried out, there are few results giving adequate accuracy. Thus, in this study, systematic experiments have been carried out in order to obtain unsteady forces acting on gas turbine cascade blades and an evaluation method for blade-exciting forces has been deduced by using experimental data. The 1st report mainly shows experimental equipments, procedure and results, and it is clarified that unsteady forces are mainly dependent with the reduced frequency.

View full abstract

Unsteady aerodynamic forces on a turbine cascade blade due to a sinusoidal gust have been studied experimentally. Turbine blades having large camber, large thickness and large turning angle, it is difficult to obtain unsteady forces theoretically. In spite of many theoretical investigations having been carried out, there are few results giving adequate accuracy. Thus, in this study, systematic experiments have been carried out in order to obtain unsteady forces acting on gas turbine cascade blades, and an evaluation method for blade-exciting forces has been deduced by using experimental data. In the 2nd report, experimental data are discussed in detail and the evaluation method is described so that the exciting forces can be obtained by using the reduced frequency and cascade conditions.

View full abstract

Sound generated by a plane jet impinging on a circular cylinder was found to be significantly reduced by making a groove in a cylinder filled with a soft material (sponge) of low acoustic admittance. The groove was of a rectangular shape and aligned parallel to the axis of the cylinder. This groove was located at a position where large-scale vortices of the jet impinge on the cylinder surface; the position of impingement was defined as the position where the r. m. s. surface pressure attained a maximum. A consideration of the mechanism of the sound reduction revealed that the depth of the groove is crucial in reducing the sound, and that the optimum depth is given by 0.6 times the exit velocity of the jet divided by the most unstable frequency of the jet. The width of the groove had only a minor role on the sound reduction. Measurement of the acoustic intensity vector field suggested that the pressure wave propagating upstream to excite the shear layer of the jet may have its origin in the pressure fluctuation produced by the separation of flow which is induced by large-scale vortices of the jet.

View full abstract

Fluid forces and moment acting on a blade of the Darrieus-type runner were measured in addition to the blade performance. The characteristics for various type of blades were evaluated from the measured data with theoretical considerations in an inviscid flow. Some features of those characteristics, including the dynamic effects caused by an unsteady relative flow around the Darrieus-blade, are shown in the present paper. For high efficiency, a noncambered thin blade with relatively long chord length is recommended. Its characteristics are high lift slope, high lift at zero angle of attack and relatively low drag in a wide no-stall region of attack angles. It should also be set tangent to the runner pitch-circle at the 50% chord point of the blade.

View full abstract

To examine the effects of a secondary flow induced by tip leakage, the velocity and pressure distributions are measured within an unshrouded centrifugal impeller with a large tip clearance. The changes of the characteristics between shrouded and unshrouded impellers are discussed in comparison with the previous reports. In spite of the considerable difference of the tip clearance, the secondary flow pattern is similar to the first report. Though a wake in the shrouded impeller of the second report is found on the suction side, the wake in the unshrouded impeller exists on the midposition of blade-to-blade near the casing because of the effects of tip leakage. The position of the largest blade loading in the unshrouded impeller changes to midpassage from first-half-passage in shrouded impeller. Therefore the absolute circumferential velocity at the exit in the unshrouded impeller is larger than that in the shrouded impeller.

View full abstract

The pumping effect of a vibrating pipe system has been studied experimentally. Relative displacement of vibrating pipe and valve, pressure and flow rate in the pumping system were measured. The relation between these time historical data is discussed and the details of the hydrodynamic mechanism of pumping effect are clarified. The energy conversion process between vibrating pipe and fluid was investigated by analyzing the fundamental pumping energy.

View full abstract

Unsteady characteristics of a hydraulic motor are examined experimentally in the speed range where negative resistance appears in a friction torque-rotational speed diagram. When the rotational speed is fluctuating around a mean value during continuous rotation of the motor, the friction torque takes a constant value and negative resistance does not appear even in the negative resistance speed range of the steady characteristic. On the other hand, in the case where the motor stops intermittently, as in the case of stick-slip or inching operation, the friction torque shows hysteresis between the accelerating and decelerating periods. The effects of the stop time, the magnitude of speed fluctuation, and the reverse of rotational speed on the unsteady friction torque are also clarified. The friction torque changes quasi-steadily while the pressure changes.

View full abstract

In the previous paper' the study of pressure waves in oil column separation led to the idea that an oil column should rupture when it is exposed to sudden tensile stress. The present paper subsequently deals with clarifying the idea of 'rupture'. In order to give evidence to the hypothesis, oil behavior in an acrylic tube during column separation is observed with a video camera and a photosensor: cavities are confirmed to emerge and disappear immediately after the reunion of separated columns. Theoretical discussion is held on a minute bubble with surface tension to see if it can consistently explain the ruptures and absolute negative pressures an oil column exhibits. In conclusion, the picture of oil rupture and cavitation is as follows: oil continuum yields to an excessive tensile force as metal does, then splits and produces a vacuum chasm into which resolved gas spreads from surrounding oil, which results in cavitation bubbles.

View full abstract

The effect of the thermal properties of the polymer coating on the dropwise condensation heat transfer was studied theoretically and experimentally. Fundamental equations describing the constriction resistance phenomena caused by the polymer coating were obtained by considering the contribution of the droplet resistance in the individual drop-size class to the thermal resistance in the so-called transient dropwise condensation. In the experiments, fluoroethylene propylene (FEP) was coated on the copper block and heat-transfer measurements were performed for dropwise condensation of steam in a pressure range from atmospheric down to 2kPa. The experimental data show that the heat-transfer coefficient is dependent upon the coating thickness. Also, the measured heat-transfer coefficients agreed satisfactorily with the theoretical predictions. It is then confirmed that the heat-transfer coefficient decreases with increasing thickness of the polymer coating due to increasing constriction resistance.

View full abstract

Based on the liquid sublayer dryout mechanism, a physical model of critical heat flux (CHF) of subcooled flow boiling was presented in a previous paper with a rather tentative correlation of a coefficient to connect the velocity of a vapor blanket with that of two-phase flow. In the present paper, a new correlation of the velocity coefficient is derived based on CHF data of various kinds of fluids, presenting a generalized method of predicting the CHF of subcooled flow boiling within the range of void fraction less than 0.7, and at a vapor/liquid density ratio greater than 0.01. Comparisons of predicted and measured CHF show fairly good accuracy up to very high degrees of subcooling.

View full abstract

Many investigations on the utilization of finned tubes have been performed to enhance heat transfer in a latent heat thermal energy storage (LHTES) systems. Detailed information on the heat transfer for the solidification process in the LHTES having tubes with longitudinal fins has been obtained experimentally and theoretically. For LHTES systems having tubes with radial fins, however, no thoretical studies to obtain the overall heat-exchanger performance appear to have been reported since usual numerical methods (e. g. a finite difference method) cannot be easily applied to the problem. In the present study, the overall performance of a LHTES system having tubes with radial fins has been obtained using a new method which the author proposed recently, and the performance has been compared with those of LHTES systems having a tube with longitudinal fins and having an unfinned tube.

View full abstract

An experimental study is made on turbulent diffusion flames stabilized by a circular cylinder in a divergent flow. In this paper, stabilization characteristics and flame structure are examined by varying the divergent angle of duct. The fuel used is a commercial grade gaseous propane supplied by exudation from a sintered metal smoothly embodied in the backward stagnation region of the rod. It is found that the positive pressure gradient greatly influences the eddy structure behind the rod, and that two different kinds of combustion patterns exist at the blowoff limit depending on the divergent angle of the duct. They are distinguished by their wake structures: one associated with Karman vortex shedding, the other without it. Also, the blowoff velocity in the former is found to be higher than that in the latter.

View full abstract

The evaporation and combustion characteristics of a single coal-water slurry droplet were studied in an elevated temperature environment and in a hot air stream by measuring the temperature of a burning droplet and observing its behavior. When the droplet was suspended in an elevated temperature environment, the decomposing combustion, that is, the combustion of an inflammable gas included in the coal, started after the dry-up. The surface reaction, that is, the combustion at the coal surface, followed after that. Thereby, the surface reaction was slightly influenced by the decomposing combustion. When the droplet was suspended in a hot air stream, the decomposing combustion started during the surface reaction process and a blue wake flame was induced. In this case, the spread of the surface reaction region was prevented by a flame of the decomposing combustion.

View full abstract

Ignition characteristics of a propane-air mixture that contains kerosene spray were studied to understand the ignition behavior of an evaporating gasoline spray injected into a cylinder of a direct injection stratified charge (DISC) engine. Ignition probability as a function of an overall equivalence ratio, a mixture ratio of propane/kerosene to the total fuel and spark energy, was measured under atmospheric conditions. Adding a small amount of kerosene spray to a lean propane-air mixture improved the ignition probability of the mixture. A lean propane-air mixture had the suitable mixture ratio of kerosene spray for ignition. The results obtained here will give basic information about the ignition behavior of an evaporating gasoline spray, because a propane-air mixture containing a kerosene spray would simulate various evaporating stages of a gasoline spray.

View full abstract

The process of flame development in a spark-ignited propane-air mixture that contained kerosene spray was observed by a high-speed VTR. Flame areas were measured on the VTR and the process of flame development was shown as a change of flame area versus time from ignition. Propagation delay, which means a period between spark ignition and the start of rapid flame development, was defined. Effects of an overall equivalence ratio, a mixture ratio of propane/kerosene to the total fuel and spark energy on the propagation delay and development rate of a flame were studied. Each lean propane-air mixture had a suitable value of the mixture ratio of kerosene spray for the minimum propagation delay and maximum development rate of a flame. When the mixture ratio of propane to the total fuel was low, the propagation delay decreased as the spark energy increased.

View full abstract

The soot formation rate at high pressures has been studied with a premixed propane-oxygen-inert gas combustion in a constant-volume vessel. The soot formation rate was measured by the laser extinction technique at the conditions of: 0.4 to 5.0 MPa for pressure, 1600 to 1900 K for temperature and 2.0 to 2.2 for the equivalence ratio. The burnt gas temperature was measured by the two-color method. It was found that the soot formation rate can be expressed by empirical pseudo first-order kinetics for surface growth even at high pressures in the present study, and that the rate constant of soot formation is only dependent on the temperature.

View full abstract

The suppression of turbulence, caused by the increase of molecular viscosity due to high temperature in flames, was investigated in two experimental approaches. First, hydrogen jet diffusion flames were examined in the wide range of Reynolds numbers of nozzle flow. Second, isothermal hot air jets were examined with various efflux conditions of the nozzle fluid, because it was supposed that hot air jets are more suitable for investigation of the effects of viscosity increase resulting from high temperature on turbulence behavior. As a result, it was revealed that the turbulence suppression takes place also in the flame of high Reynolds number and exerts an important influence on the flame structure. The hot air jets showed a development of jet boundary layer similar to jet diffusion flames in a certain condition. This result suggests that, if a pertinent turbulence model can be obtained to represent the flows in the hot air jets, it may also be useful in the modeling of jet diffusion flames.

View full abstract

Numerical computations were made of evaporating and nonevaporating transient sprays based on a Lagrangian equation of drop motion and Eulerian conservation equations for the gas. Drop dispersion and penetration, velocity correlation of drops and gas, drop radius distribution, evaporation of drops, fuel vapor penetration and so on were predicted to clarify the spray configurations and characteristics. Comparisons between computations and experiments were made to illustrate the availability of the prediction procedures.

View full abstract

In order to elucidate the disintegrating mechanisms of a diesel spray, the breakup length was measured by an electric resistance method. A reattachment distance of separated liquid flow in a nozzle and a spray were observed photographically. The mechanism of the two disintegrating processes of the jet was investigated by separating an ambient effect from a disturbed effect of the nozzle. The atomizing mechanism of the jet was governed by both the initial disturbance due to the nozzle geometry and an interfacial force between the jet and the environment. If a strong disturbance was given to the internal flow in a nozzle, the jet would be broken up into a spray flow.

View full abstract

A three-dimensional computational analysis program of gas-liquid two-phase flow is developed in order to simulate fuel flow in the intake menifold of an automobile fuel supplier. This program considers physical models as follows: (a) movement of fuel droplets and their deposition on the manifold wall; (b) temperuture change and evaporation of the droplets: (c) evaporation of fuel deposited on the wall: (d) transport of fuel vapor evaporated from the wall: and (e) force of the droplets' influence upon air. Calculated results show that all fuel droplets injected from the fuel injector deposit on the wall, and when the wall temperatures are 20°C and 80°C, 95% and 37% of the injected fuel flow into the cylinders in the state of liquid film, respectively. The remaining fuel flows into the cylinders as vapor.

View full abstract

To meet the demand for the futuristic fuels, the authors have been studied an ethanol-stratified charge engine. The test was carried out using an aircooled single cylinder 4-cycle engine having a two-hole diesel-type injector, full-transistor-type ignition system and an ignition chamber. The control parameters of this combustion system are the injection direction to the swirl and the relative position of the ignition plug to the fuel spray in the ignition chamber. The tests show that injection against the swirl in the combustion chamber is good for the combustion stability for a lisht load, and central injection to the combustion chamber is best for power and thermal efficiency for a high load. Also, the effect of the injection, ignition timing, ignition energy and intake air throttling at light load have been studied in detail.

View full abstract

Seven optical fibers are introduced into the S. I. engine in order to detect the position of autoignition occurrence preceding knocking. Light signals through the fiber are synchronized with the autoignition photographs taken by the high-speed shadowgraph technique, and pressure signals. The location of the fiber which indicates the strongest light intensity after flame propagation corresponds to the autoignition location judged by the shadowgraph pictures and supposed by 4 pressure signals.

View full abstract