Unsteady sprays are used in practice in both DI and IDI diesel engines. The former engine is of great advantage in thermal efficiency to the latter one. However, it is unavoidable that sprays impinge surely on a wall of the piston cavity. The effect of impingement on the mixture formation in engines plays a significant role in the improvement of the thermal efficiency. This article summarizes recent important papers on the characteristics, that is, the external shape, the adhesion of fuel on the wall and the internal structure, of unsteady spray impinging on the wall.

View full abstract

Effects of both the dimeter ratio D/d and the aspect ratio L/d of a pair of circular end plates on the spanwise characteristics of flow around a circular cylinder were examined experimentally at a Reynolds number of 3000. D/d, which indicates the ratio of end-plate diameter to cylinder diameter, was altered from 4 to 15, and L/d, which denotes the effective aspect ratio of circular cylinder given by the interval between two end plates, was in a range below 60. The experimental results reveal the following : The base pressure of the cylinder with end plates of a suitable diameter ratio becomes constant in a wide region of cylinder span between two end plates. The spanwise range in which the base pressure is uniform is smaller in the case of diameter ratios D/d=4 and 15 than in the case of a suitable ratio. The base pressure coefficients in the center of span of all the cylinders with end plates of D/d=4∼15 become constant for aspect ratios beyond L/d=30. Furthermore, the base pressure coefficient of the cylinder with end plates of the most suitable diameter ratio D/d=8, becomes constant for small aspect ratios until about L/d=8. The minimum limit of aspect ratios in which the Strouhal number in the center of span maintains a value of 0.21 is smaller in the case of D/d=8 than in the cases of other diameter ratios.

View full abstract

If a three-dimensional, turbulent, round wall jet is issued tangentially along a circular cylinder, the jet scarcely deflects and does not attach to it since the jet comes into contact with a small area on the cylinder surface near the nozzle exit. However, when the jet is issued at an angle, the jet flows attach to it at the nozzle exit. It is a very important to elucidate the flow characteristics, because they are concerned with enhancement of wing lift, control of boundary layer, film cooling, and so on. In this study, the effects of ejection angle on the flow characteristics, velocity and pressure distributions, decay of maximum velocity, spread of jet, separation (attachment distance), and so on, are clarified experimentally.

View full abstract

The oscillation characteristics of water columns in a water-type Stirling engine, called the Fluidyne engine, are investigated experimentally and theoretically in both states of damped free and forced oscillations. First, the equation of motion is solved by taking into account the quasi-steady friction losses and external pressure, which enables us to estimate the oscillation characteristics of the water columns. Then the experimental study is conducted to monitor the water column oscillation in the tube. The results suggest that theoretical analysis can estimate well the loading characteristics in these machines.

View full abstract

Thin-film flow on a spinning doughnut-shaped substrate was analysed numerically. A simple one-dimensional non-Newtonian flow model was employed to calculate the film flow. The time-dependent change in the rheology of the film liquid, due to the evaporation of solvents in the liquid, which is spatially varying, was considered in the calculation. According to the numerical experiments, the spatial distribution of the solvents' evaporation rate, as well as the liquid rheology, showed a significant effect on the final thickness distribution of the film. On the other hand, the initial thickness distribution of the film did not show any effect on the final thickness distribution.

View full abstract

The one-dimensional two-fluid energy and momentum equations have been developed to predict the axial distribution of liquid level or void fraction in steady cocurrent gas-liquid stratified flow in horizontal circular and rectangular channels. Two kinds of critical depth derived from the energy equation are adopted as a boundary condition at the channel exit depending upon whether the liquid level in the exit tank is kept above or below the channel. The predicted results are compared with the experimental data of other investigators as well as the present authors. Good agreement was obtained for flows with a smooth gas-liquid interface.

View full abstract

Fully developed flow in strongly curved tubes was numerically solved by SIMPLER method. The physiological flow condition was performed by systolic period with sinusoidal volumetric flow rate and following diastolic period. Numerical calculation was carried out under the following conditions, Dean number x=393, frequency parameter a=4∼27, curvature ratio δ=1/2, 1/3 and 1/7, and intermittent parameter η=0∼1/2, where η is the ratio of a systolic time to a whole period. Blood flow is generally characterized by intermittent flow having diastolic period, but the flow does not diminish completely during the diastolic period and retains some strength until the starting phase of the next systolic period. However, the axial flow profiles in systolic period become close to the typical profiles appearing in oscillatory flow with high a. The secondary flow is strongly affected by the conditions at the starting phase, so η plays a major role in the resulting vortex pattern. During the systolic period, the axial and secondary flows gradually decrease as a whole. In detail, reverse axial flow appears in the circumferential region of the tube at the entering phase of the systolic period as behavior of stop flow, and at the same time the development of a boundary layer restarts in this region. This process induces the secondary flow. In the case of small η, the secondary flow forms an additional vortex near the inner corner, while in the case of large η, vortex formation is not seen but the strength of secondary flow increases locally. The strength of the flow does not decrease monotonously. Aortic arch flow strongly depends on the flow from the precycle. The intermittent flow in a curved tube is strongly affected by the flow in the stationary diastolic period.

View full abstract

In the design of the local exhaust ventilation systems of high performance, it is of importance not only to maintain the center-line velocity as high as possible at a certain distance from the hood opening, but also to widen the high-velocity area in the direction perpendicular to the center-line for the respective contaminant sources. To meet these demands, the effects of hood geometry on the suction performance are studied in the three-dimensional flow field for freestanding circular, square and rectangular hoods. The nondimensional center-line velocity with respect to the opening is a similar distribution irrespective of hood shape and it agrees with Dallavalle's empirical formula. To elucidate the characteristics of the hood geometry, the velocity is normalized by the take-off, In this case, the center-line velocity distributions agree with each other only for the same opening ratio, irrespective of the cross-sectional geometry. The flow fields with respect to the opening in the horizontal and vertical planes including the hood axis show different velocity contours except in the case of hoods with the same flare angle. From the results of velocity contours with respect to the takeoff, the predominant regions are defined for each hood with a different area ratio. From this result, the optimum hood opening ratio can be selected in accordance with the respective sources. The flare angle of the minimum head loss, regardless of the hood geometry, was also found.

View full abstract

To make clear the behavior of cavitation erosion, flow aspects and bubble-collapse pressures of Karman-vortex cavitation behind several symmetrical wedges are investigated under various flow conditions in due consideration of wall effect. The time histories and intensity distributions of the pressures are measured by means of a pressure transducer and pressure-sensitive films. The cavitation aspects are observed by high-speed photography. The occurrence of bubble-collapse pressures is synchronized to the Karman-vortex shedding frequency. The pressures are most intensive at about cavitation number σ=1.0, where is the transition region between subcavitation and supercavitation. For an increase in wedge angle or angle of attack, the intensity distributions of the pressures distinctly extend towards downstream.

View full abstract

As a simple model of the large-scale structure in the laminar boundary layer, the devolopment of multi-disturbances is calculated using a three-dimensional discrete vortex filament method. Two equivalent initial disturbances are applied to the same filament and are stretched by the streamwise velocity. The top regions of the disturbance are deformed by the induced velocity. As a result, the velocity profile of the boundary layer is also deformed as the disturbances develop and the initial disturbances are stretched in the downstream direction. Then the hair-pin-shaped structure deforms and pinches out. A unique vortex loop is created and the sweep structure appears in the vicinity of the wall which does not appear in the case of deformation of a unique disturbance.

View full abstract

A numerical analysis is performed for a rotary pump using the Weis-Fogh mechanism, which is a lift-generation mechanism in the hovering flight of small insects. Two-dimensional flow past a rotor in a square channel is investigated by the discrete vortex method. Distribution of vortices shed from each wing of the rotor and velocity vector fields are studied. Some patterns of forces on a wing of the rotor are also shown. As the ratio of the moving speed of the wings to the uniform flow velocity increases, the pressure rise behind the rotor increases. When the ratio is small, this model works as a water turbine. Results of this analysis are compared with those of the experiments, and the characteristics which differ from those of the reciprocal type are also considered.

View full abstract

The vortex model of the blob technique, which is one of the previously proposed vortex methods, was modified to take account of the viscous effect correctly with a simpler procedure of calculation. To examine the usefulness of the modified technique, the interaction of two vortex rings is calculated with it and two other vortex methods : the vortex filament technique and the vorton technique. As a result of the comparative calculation, it was confirmed that not only the process of connecting the two rings into a single ring at an early stage of interaction, but also the process of splitting the single ring into two rings at the following stage is reasonably simulated by the modified technique.

View full abstract

When the change rate of the stress tensor is approximated by a time derivative, the constitutive equation is reduced to the linear function of a deformation rate tensor. The derived pseudo-viscosity decreases as relaxation time increases. To enhance the pseudo-viscosity matrix, a relaxation method is also presented. By the proposed formulation, the stress overshoot effect of a transient Couette flow is definitively simulated.

View full abstract

A numerical method using boundary-fitted curvilinear coordinates (BFC) is applied to compute non-Newtonian flow equations in primitive variable formulations for an incompressible inelastic power-law fluid. The flow equations are solved on a special staggered grid by the method of lines using the rational Runge-Kutta time integration scheme and the central finite difference method for spatial discretization, while the Poisson equation is solved by means of the successive overrelaxation (SOR) method using the modified simplified marker and cell (MSMAC) method. It is concluded that the results using BFC are comparable to those using Cartesian coordinates, and the proposed method can be applied to the numerical solution of flow problems of a power-law fluid in arbitrary computational domains.

View full abstract

A numerical technique for investigating the mixing characteristics of supersonic turbulent flow with different gas species, in which a two-equation turbulence model was employed, was presented in the first report. In the previous report, uniform inlet flow profiles are assumed for both high-speed and low-speed side gases. In this report, in order to clarify the effect of inlet flow profiles on the mixing phenomena, distributed flow profiles obtained for a supersonic wedge-type nozzle outlet are adopted as the inlet boundary conditions of mixing calculation. The general conclusion is that the values of mixing layer thickness with distributed inlet flow profiles are wider than those with uniform profiles. Also the growth rates of mixing layers with distributed inlet flow profiles plotted against the change of velocity ratio, and density ratio have a similar tendency to those with uniform profiles.

View full abstract

In order to increase the lift capability and/or to reduce take-off speed of a wing-in-ground effect craft, it is necessary to increase the lift coefficient of the wing in the ground effect. For this purpose, screening tests were conducted by applying various methods of lift enhancement to a ground-effect wing model. A lift coefficient of nealy 2 was obtained with a slotted airfoil system consisting of three slots and four airfoil elements. Longitudinal stability was recognized as an important factor.

View full abstract

In order to increase the lift capability and/or to reduce take-off speed of a wing-in-ground effect craft, it is necessary to increase the lift coefficient of the wing in the ground effect. For this purpose, experimental investigation has been made to evaluate the power augmented Ram wing with forward-mounted propulsion through a low-speed wind tunnel. A lift coefficient above 3 was obtained by the powered ground effect. However, It was recognized that longitudinal stability should be addressed.

View full abstract

In this paper, we discuss the Fourier transform method (FTM) for particle image velocimetry. We point out that the conventional Fourier integrand may diverge for fluctuating velocity fields, and give incorrect results. We propose several improvements for FTM that suppress the divergence; these are based on spatial frequency filters, normalization or their combination. Two types of numerical simulations are carried out to examine the performance of the proposed methods. One deals with fluctuating velocity fields by adding white noise to a uniform flow; the other deals with a quasiunsteady wake past a circular cylinder given by a finite difference method. It is found that the R. M. S. error can be reduced to about 1.5 pixel by using a proposed FTM that is based on normalization and emphasis of low-frequency information.

View full abstract

In order to measure the transient density distribution of the exchange process caused by a negatively buoyant jet, an automatic and high-reliability system has been developed. The upward jet of air into a box with helium was visualized using the Mach-Zehnder interferogram. The VTR images of the Mach-Zehnder interferogram are automatically analyzed by an engineering work station, resulting in the desired two-dimensional transient density distribution. The stratification of air and helium was clarified from the measured density distributions.

View full abstract

A method to determine the wide variation of hydraulic losses with the swirl velocity at the outlet of an adjustable blade runner is presented by analyzing the performance curves of a bulb turbine under the shock-free inflow conditions. The variation of the losses and the pressure recovery coefficients of the draft tube with the swirl velocity is investigated with two model bulb turbines having discharge specific speeds n_SQ of 158 and 171. It becomes clear that the optimum swirl flow angle at the runner outlet under the minimum hydraulic loss is approximately nine degrees as measured from the axial direction, and the variation of the losses can be determined by only the swirl flow angle at the root-mean-square radius of the runner outlet.

View full abstract

A Shape-controlled airfoil was developed by use of shape memory alloys (SMA). Two-way change in the blade shape was realized by use of a differential two-way element in which the two different shapes were memorized. The developed airfoil was tested in the wind tunnel in order to check the effect of the shape change on the characteristics of the airfoil. Flow visualization experiments in a smoke tunnel as well as the traverse of the wake behind the airfoil showed that the shape change by electrically heated SMA gives a marked change in flow around the airfoil near the stall angle of the original shape. As the result of this study, it was found that the developed SMA actuator is effective for the control of flow separation from the blade surface.

View full abstract

Measurements were made of the departure frequency of coalesced bubbles from horizontal disks, of which diameter ranges from 2mm to 30mm. In a sintered metal disk nitrogen gas was fed steadily into the disk in liquid (water, ethanol or freon-113) and passes out as bubbles through the disk. The departure frequency of bubbles was also measured for the boiling of water, ethanol and freon-113 on thin horizontal wires of 0.2mm∼3mm in diameter. Bubble frequency was determined under atmospheric pressure by a video camera and a high speed camera. Semi-empirical correlations of frequency were proposed on the basis of the dimensional analysis of a momentum equation of bubbles and the results of measured departure frequency.

View full abstract

Macrolayer thickness was determined on the basis of the relationship q_CHF=δ₁ρ₁H_fgf, which was derived from the dry-out model of liquid film formed repeatedly on heated surfaces. Critical heat flux q_CHF is measured on a horizontal heated disk of 20mm in diameter in the range of pressures from 0.02 MPa to 0.4 MPa for water and ethanol. Departure frequency of large coalesced bubbles, which entirely cover the heated disk, was calculated from the relationship f=0.215{g(ρ₁-ρ_v)/ρ₁}^5/9/(ν₁·D³)^1/9. Theoretical relationships correlating the obtained data of macrolayer thickness are derived by the dimensional analysis, under the assumptions that a macrolayer is formed by the coalescence of primary bubbles or coalesced bubbles. The macrolayer data are well correlated with the derived correlations.

View full abstract

The correlation of critical heat flux (CHF) is derived for thin horizontal wires on the basis of the relation, q_CHF=δ₁ρ₁H_fgf. This relation was derived from the model, in which liquid film on the surface of wires is just dryout when coalesced bubbles depart from the surface. The correlations of δ₁ and f are given in the third and forth papers of this series, respectively. The macrolayer correlation on flat surfaces is modified for very thin wires by taking into account of the effect caused by the thin diameter of heated wires. Calculations of CHF are made as a parameter of wire diameter for various liquids. The calculted results are compared with the values measured by Sun et al. and the present authors, and the excellent agreement between them are obtained.

View full abstract

Critical heat flux (CHF) on vertical wires was measured for water and ethanol under atmospheric pressure. The diameter of wires ranged from 0.3mm to 3mm and their length from 4mm to 200mm. For shorter wires, CHF occurred at the top and increased with decreasing length, while for longer wires, CHF occurred in a lower region of the wires and did not vary with diameter. A physical model for CHF on vertical wires was constructed based on a CHF mechanism, in which CHF is caused by the dry-out of liquid macrolayers formed on a heating surface. In the model it is assumed that coalesced bubbles moving upward along a wire have a liquid macrolayer on their bottom, and the motion of the coalesced bubbles is determined by the balance of force between buoyancy and drag. CHF on the longer wires occured where coalesced bubbles begin a zig-zag or spiral motion. The predicted CHF and the location of CHF were compared with experimental data and fairly good agreement was obtained.

View full abstract

Experimental results are presented on the variations of liquid film thickness, liquid film flow rate and film flow direction on the wall along the axis and the circumference of a tube in an air-water mixture flowing upward in an annular mist flow regime through a C-shaped bend of 144-mm radius of curvature with a 24-mm-inside-diameter tube lying in a vertical plane. In order to study the effect of gravity on flow direction, some experiments are also carried out on downward and horizontal flows. in the bend. The flow of liquid film is greatly influenced by the secondary flow caused by centrifugal force within the gas core and gravity, and exhibits some notable features.

View full abstract

In the previous report, we have studied an anisotropy of the thermal diffusivity of molded PMMA spiral-flow samples by means of the forced Rayleigh scattering method. This anisotropy of thermal properties was caused by the orientation of polymer molecules under shear flow field in the mold. In the present paper, the anisotropy of the thermal diffusivity due to molecular orientation in shear flow was experimentally measured for the sample in molten and flowing condition. One-dimensional flow of molten polymer was created in a thin rectangular visualized conduit with a plunger extruder, and the anisotropic behavior was investigated for molten polystylene at different temperatures and different velocity profiles. It was shown as possible to measure the thermal diffusivity of molton polymer in its flowing condition by mean of the forced Rayleigh scattering method. The ratio of a∥, the thermal diffusivity in the direction parallel to the flow, to a ⊥, that rectangular to the flow, increases with a mean velocity increase. It was shown that all of measured thermal diffusivity can be correlated using Re number as a parameter.

View full abstract

The melting characteristics of liquid ice in a rectangular cavity were studied experimentally. The liquid ice, a mixture of ice particles and ethylene-glycol aqueous solution, was heated by one of the vertical walls of the cavity. The shape of the mush-liquid interface, melting rate, and local/mean heat-transfer coefficient at the heated vertical wall were observed and measured under a variety of conditions of heat flux and initial concentration of the aqueous binary solution. It was found that the formation and growth of double-diffusive layers due to the thermal and solutal buoyancy forces have a great influence on the melting process of the liquid ice.

View full abstract

A correlation for the boiling heat transfer coefficient in the heating section of a two-phase thermosyphon heat pipe is developed in comparison with experimental data. The experiments are conducted for a vertically oriented thermosyphon pipe using three kinds of working fluid, water, Freon R113, and ethanol, at a wide pressure range of 0.1 to 20 bar and a fill charge rate of 0.3 to 0.9. The correlation is obtained by modification of Kutateladze's nucleate boiling correlation in terms of an additional two parameters, the density ratio of liquid and vapor and the surface roughness. The correlation is shown to be in good agreement with the experimental data at a system pressure above 1 bar. The experiments also indicate that the adiabatic wall temperature is less than the saturation temperature corresponding to the system pressure in the pipe.

View full abstract

Temperature distribution in a layer of an active thermal insulation system was measured experimentally. A semitransparent porous layer was heated by a gas burner, and air was injected from the back face of the layer. The temperature in the layer was measured by thermocouples. The temperature distributions were compared with numerical solutions. The thermal penetration depth of the active thermal insulation layer with gas injection can be reduced to 3mm. The surface temperature of a conventional insulation layer without gas injection reached 1500K, and the temperature at the back surface of a 10mm thick layer reached 600K. The transient temperature of the active thermal insulation reached a steady state very quickly compared with conventional insulation. These characteristics agreed qualitatively with numerical solutions.

View full abstract

Melting characteristics of a solid phase-change material in an inclined rectangulr heat-storage enclosure heated from one side are investigated numerically. It is found that a two-dimensional model simulates the melting behavior of the inclined rectangular heat storage, as compared with the results of a one-dimensional model. In each quasi-static step, steady-state natural convection in the liquid phase is calculated by directly solving the governing equation of motion with a finite difference technique. It is noted that the time development of the melting process depends strongly on the natural convection behavior, which is changed by the inclination angle and geometrical dimensions of the rectangular heat-storage enclosure.

View full abstract

With modifications of a two-equation heat-transfer model, a new approach to analysis of heat transport in free turbulent shear flows is proposed. The modified model was tested by application to different homogeneous turbulent flows, namely decaying isotropic turbulence, homogeneous nonisotropic turbulence under a return-to-isotropy process, and homogeneous turbulence with a simple shear. To represent the nonisotropic nature of heat transfer in free turbulence, we have modeled the eddy diffusivity for heat by a second rank tensor on the basis of the generalized gradient diffusion hypothesis. At the same time, when treating the time scale which characterizes the transport process of a passive scalar, we have introduced the scalar-field time scale τ_θ=((θ²)^^^-/2)/ε_θ. The predictions of turbulent heat flux, temperature variance and its dissipation rate are all in satisfactory agreement with the experimental results.

View full abstract

In the present study, a flow field developed in an opposed jet burner and its turbulence characteristics were measured in detail by a hot wire anemometer. The flame structure was elucidated by schlieren photographs and instantaneous temperature measurements with a thermocouple. Extremely strong turbulence of small Scale is generated by impingement of two mixture flows. For the case of low mixture supply velocity, a flame within the transition regime from wrinkled laminar flame to distributed reaction zone is produced in the opposed jet burner. With increasing mixture supply velocity, the structure of the flame zone changes to the distributed reaction zone. These two types of structure correspond to region 2 and region 3 of the 3-region model of turbulent premixed flame structure.

View full abstract

The present authors previously proposed a modification of the k-ε turbulence model in the modelling of jet diffusion flames to represent the so-called laminarization phenomenon, caused by combustion in the low-turbulence region, and recently found, in experiments on diffusion flames stabilized on a bluff body, that this phenomenon also exerts an important influence on the structure of these flames with recirculation zone. In the present study, the above modified k-ε model was applied to the modelling of this bluff-body diffusion flame in order to examine applicability of the modification and the ability of the k-ε model to represent such a complex flow field. As a result of comparison of the calculated results to experimental ones, it was found that the consideration of the laminarization phenomenon enables the k-ε model to provide preditions of encouraging quality, though their quantitative discrepancies are still discernible.

View full abstract

The purpose of the present study is to investigate the possibility of the use of the N₂ plasma De NO_x process for combustion exhaust gas. Removal characteristics of nitrogen monoxide are studied under the Zeldovich mechanism and calculated by the use of the computer program for chemical equilibrium theory. The temperature and velocity distributions of plasma wake were measured along the jet axis. The contamination of plasma flow due to the erosion of the electrodes was observed by plasma spectroscopy. Optimum flow rates of plasma N₂ gas and simulated exhaust gas are determined after consideration of stable plasma operation and the formation of N atoms in plasma.

View full abstract

The purpose of this study is to investigate the possibility of removing NO_x contained in combustion exhaust gases by N₂ plasma injection. The simulated exhaust gas (N₂+O₂+CO₂+NO) was mixed with nitrogen atoms produced in the N₂ plasma, and the effects of O₂ and CO₂ on the reduction of NO_x and CO generation were studied experimentally. The NO_x removal ratio η of 40 to 50% is observed under the conditions of O₂ concentration, 3 to 5%, plasma power, 600W, and CO concentration contained in gas, about 300ppm. Removal ratio η is highest for a low reaction temperature less than 500K. The value of η is reduced to 20% due to the addition of H₂O to gas. However, the recovery of η is obtained by changing the gas injection site to plasma; this is caused by the decrease of OH radical formation. Retention time required for reaction is on the order of several tens of milliseconds.

View full abstract

To elucidate the atomization process of a liquid jet across a high-speed airstream, the spray shape, the mass flow rate per unit area and the Sauter mean diameter were measured. Under an airstream velocity of 140m/s or 70m/s and injection velocity of 10m/s, 20m/s or 40m/s, the Sauter mean diameter and the maximum liquid mass flow rate per unit area in the spray hardly changed with the liquid injection velocity. However, with increasing air velocity, the Sauter mean diameter decreased and the maximum liquid mass flow rate per unit area increased. Rearranged spatial distribution of the liquid mass flow rate per unit area in a horizontal direction agreed with the profile by the Gauss's function. An equation for width of spray was obtained from the photographic results.

View full abstract

To elucidate the atomization process of a liquid jet across a high-speed airstream, the spatial distribution of liquid, drop diameter and drop trajectory were calculated. In this model, a liquid column is not incorporated. Drops ejected from the injector have the same velocity V_l but different drop diameters, which are defined by the volume distribution of drop size. The ejected drop is broken up by the airstream. Calculated spatial distribution of liquid agrees well with measured spatial distribution of liquid. The trend of the calculated drop diameters agreed well with the trend of the measured Sauter mean diameter. The drop diameter decreases rapidly, near the injector especially along the outer line of the spray.

View full abstract

The atomization characteristics of intermittent-impinging-type nozzle was investigated. Fuel supplied from the nozzle were impinged on each other. The interval time of the liquid supply was constant at 10ms and the supply pulse width was changed from 1ms to 10ms. The mean diameter and velocity were measured by a phase-doppler-type particle analyzer. Effects of impingement angle, liquid pressure and supply pulse width on droplet size were clarified. When fuel was injected intermittently, large droplets were generated at the beginning of fuel supply. The construction of a fuel using impinging atomization was also clarified.

View full abstract

Inproving the combustion of spark ignition engines under the conditions of medium or low speed and low load contributes largely to the decrease of fuel consumption under the city driving conditions. As one of the utilities of this simulation model, the variable shroud mechanism to increase swirl velocity is proposed to improve combustion. This mechanism is mounted at the intake port near the intake valve to control swirl velocity as well as the amount of fuel air mixture. In this paper, by using the predictions of the simplified simulation model, it is confirmed that this mechanism is effective in considerably increasing swirl velocity at low load and low speed conditions and furthermore, to reduce the fraction of burned gas in the fresh mixture charge.

View full abstract

Modeling of turbulence is one of the most important issues in calculating the in-cylinder flow of internal combustion engines. The authors have been engaged in evaluating the applicability of the k-ε turbulence model and also developing modified k-ε models, such as the C_D revised k-ε model which has the ability to reproduce turbulent characteristics in nearly isotropic fields, and another modified k-ε model which shows extensive multiplicity of predicting the mean velocity on any kind of turbulent field. This basic study has been made on axisymmertrical flow frields. In this report, detailed comparisons between the calculation and measurement are made to evaluate these turbulence models in an axisymmetrical field with a bowl-in piston. Also, a newly designed 3-D cylinder head that employs a variable swirl-intensity mechanism with a flat piston configuration is built for the test. Obtained conclusions in the 3-D case are fundamentally the same as those derived from the axisymmetrical cases.

View full abstract

A new finite-element method is proposed for solving unsteady incompressible Navier Stokes equations. The basis of the method is to shift the governing equations to the upstream direction by a distance proportional to the local flow velocity multiplied by the time increment. The method thus differs from most conventional upwind formulations which shift the governing equations based on the element dimension. The proposed method guarantees reasonable stability for simulations of high Reynolds number flows without introducing excessive numerical diffusion which is often encountered with conventional upwind methods. The accuracy of the proposed upwind method is first demonstrated for a simple advection-diffusion problem known as 'rotating cone'. Several simulations of unsteady laminar flows are then performed to test the full Navier-Stokes formulation. The simulated results show good agreement with the experimental data. Hence, the proposed method seems quite promising for simulations of unsteady flows in complicated geometries.

View full abstract

A numerical solution for an axisymmetric non-Newtonian flow of an incompressible inelastic power-law fluid using a primitive variable formulation is presented. The present method is based on the method of lines using the rational Runge-Kutta time integration scheme combined with the central finite difference method for the spatial discretization. The Poisson equation is solved by means of the SOR method using the concept of the SMAC method. As a test problem, numerical results for a circular pipe flow with uniform inlet velocity are shown and compared with the analytic fully developed velocity profile to test the validity of the present method. Next, numerical results for the flow through a pipe with abrupt contraction are shown. The difference between Newtonian (n=1) and non-Newtonian fluids (n=0.75 and 1.25) for the Reynolds number (Re) from 0.01 to 100 is discussed concerning the flow pattern. It is evident that the size of the secondary vortex increases with increasing volumes of n. The entry length and the excess pressure drop in the contraction increase with decreasing values of n. It is confirmed that the present method is suitable for the numerical solution of a power-law fluid.

View full abstract

The principle of thermal conductivity measurements as illustrated by the transient hot wire method is based on the assumption that thermal properties of a medium are constant, independent of temperature. However, since the thermal conductivity of the medium usually changes with temperature, the temperature rise during the measurement is a source of error. In this study, the theory of thermal conductivity measurements for such a nonlinear medium using the transient hot wire method is investigated analytically, and is compared with conventional theory for a constant conductivity medium. Furthermore, experiments are performed for a medium with a comparatively large temperature coefficient of thermal conductivity and are compared with the theory for the nonlinear medium.

View full abstract