Mean flow and statistical property measurements were made as a function of the longitudinal position with the rib cycle in the near-wall region for a boundary layer over a d-type rough wall at R_θ≒800. The experiments showed that both profiles of the mean velocity and the turbulence intensity varied to be wavy in the similitude of a repeated-rib geometry, and that both quantities markedly increased at the location near the upstream shoulder of the groove. These profiles are completely different from those of the d-type rough wall flow at R_o = 5000. A discussion of some terms in both the momentum and the turbulent energy equations is given in order to describe the feature of present flow near a repeated roughness geometry.

View full abstract

In the 1st report, a theoretical and experimental analysis of flow patterns occurring around a single flat plate set on the lower wall of the flow passage, as the simplest type of turbulence promoter, was described. In the 2nd report, an experimental analysis of the case of a single flat plate set on the upper wall of the flow passage was described. This 3rd report describes an experimental analysis of the case of many flat plates set on the lower wall of the flow passage. The relationships between the coefficient of loss ζ and the Reynolds number Re, the number of plates N, the pitch of plates P are examined. A new nondimensional number S is proposed as a measure of the effect of the duct wall and the arrangement of many flat plates, and how the coefficient of loss ζ, can be expressed as a function of the number S is explained. These experimental results are very useful not only for the design of turbulence promoters but also for the design of heat exchangers.

View full abstract

The stability of flow in the narrow spacing between parallel conical surfaces rotating in the same or opposite directions is investigated by means of a linear stability analysis. The critical Reynolds number is obtained for semivertex angles, θ, from 1°to 90°and for ratios of angular velocities, μ, from -1.75 to 0.75. Four different kinds of instabilities are found to occur at certain combinations of θ and μ. Characteristics of instabilities are clarified for various values of θ and μ.

View full abstract

Strouhal number and base pressure coefficient of a cylinder with an elongated rectangular cross section of a width-to-height ratio (the B/H ratio) of 4 to 15 were measured in the range of Reynolds numbers between 200 and 7×10³ in a wind tunnel, and flow around the cylinder was visualized by means of the electrolytic-precipitation method in a water tank. It was found that there exist two components with different Strouhal numbers in a flow field around the elongated cylinder with the B/H ratio of 5.5 to 6.5 in the limited range of Reynolds numbers of 600 to 2×10³. It is clarified by experiments on a model whose nose is shaped like an aerofoil that the component with a high Strouhal frequency is induced by the vortices shed from trailing edges. The low Strouhal component is due to the oscillation of flow over the side walls accompanied by the movements of separation bubbles, so its value of Strouhal number, which is based on the streamwise length of the cylinder, remains constant; however, with further increase in the B/H ratio, it increases in a stepwise manner to values that are equal to the integral multiples.

View full abstract

It is the purpose of this paper to point out that the classification of column separations in a horizontal water pipeline as an upstream-type, a midstream-type and a downstream-type is very useful for investigating column separations and analytical methods. Detailed experiments on column separations were conducted using an acrylic pipeline which was 200 m in length and 15.2 mm in diameter. Simulations of column separations by representative analytical methods were carried out. From comparisons of experimental results and calculated results, according to the classification, characteristics of column separations and analytical methods were identified very clearly.

View full abstract

The authors studied the instability of the meniscus of a viscous liquid between two closely spaced tilted plates when they are separated from each other in the normal direction of the plates. The instability has a close relation with that of the creeping flow in a Hele-Shaw cell and is caused by viscous forces In the present paper, a local capillary number is newly defined by the local Velocity of the leading meniscus the instability criterion is theoretically deduced. The simple criterion for the instability concerned with the local capillary number and the nondimen sional wave number for perturbation on the meniscus agrees well the experimental results. The theoretical formulation of the disturbances of the free surface predicts fairly well the experimental wavelength.

View full abstract

A new method of numelical analysis of a viscoelastic fluid flow is presented. The present method is derived by applying Lax's scheme to the constitutive equation and makes it possible to deal with a flow at a high Weissenberg number and a finite Reynolds number. The present numerical results are compared with experimental observations of Cochrane et al. on a flow over a protuberance, and the numerical predictions of a vortex behind a protuberance agree with the experimental reaults. A planar contraction flow is simulated numerically by the present method and the behaviour of the secondary flow vortex is investigated in detail. As a result, three pattelns of entry flow are observed.

View full abstract

A new zonal approach for compressible viscous flow computations using a TVD finite volume method has been developed. The two-dimensional, Reynolds-averaged Navier-Stokes equations are discretized spatially by the finite volume formulation. The inviscid fluxes at the cell interfaces are evaluated through the MUSCL-type approach of an upwind scheme. In the present approach, the computational domain is divided into non-overlapping zones. The zonal boundaries are constructed from the cell interfaces, since the finite volume formulation with the cell-centered control volumes is used. Consequently communication from one zone to another is accomplished by the numerical fluxes through the zonal boundaries. The use of the finite volume formulation can ensure the unity of zonal boundaries and the complete conservation of numerical fluxes at the zonal boundaries, which results in a high-accuracy zonal approach. In order to demonstrate the versatility of the present zonal approach, numerical results are presented for viscous flows through a transonic turbine cascade.

View full abstract

Flow behavior of rotating stall in a linear cascade of NACA 65 (CA30) 10 blades with a solidity of 1.0 and stagger angle of 30 degrees, is studied numerically. The numerical method which was developed for two-dimensional incompressible separated flows of high Reynolds number around an isolated blade has been extended and used for cascades of blades. Flow behavior and timewise variation of mass flow through each blade passage clearly show the propagative features of rotating stall for a wide range of inlet flow angles. The phenomenon of stall vortex shedding, which is similar to that around isolated blades, is observed to occur in the stall cells. The physical characteristics of rotating stall seem to be well understood in terms of the features of these vortices. The total pressure-loss performance of the cascade is examined for both stalled and unstalled flow regimes. The static pressure and the velocity distributions ahead of the cascade are compared with the features of the actuator disc analysis.

View full abstract

Hot-wire measurements of the near-wall velocity profile in oscillating turbulent channel flows were performed at the intermediate frequency range, where wall shear, inertia and pressure gradient are all important but the frequency is much smaller than the bursting frequency. Viscous sublayer thickness, wall shear stress and the phase averaged wall law were examined. The phase averaged velocity profile in the near-wall region was found to obey the steady-state wall law if the wall coordinate was defined using the phase-averaged friction velocity, although the amplitude of the velocity oscillation changes from the laminar-Stokes profile to the quasi-steady turbulent profile with increasing non-dimensional Stokes thickness.

View full abstract

It is known that low-cycle system oscillation occurs in piping systems which are used with cavitating high-speed pumps. Most available papers on this subject attempt to predict this occurrence through linear transfer matrices of pumps under cavitation both experimentally and theoretically. There are few papers, however, that discuss low-cycle system oscillation with the oscillation cavitation occurring in the reverse flow at the pump inlet. The following discusses three types of low-cycle system oscillation defined by synthetic studies and the effects of NPSH, pump flow rate, rotational speed and pipe length are clarified experimentally.

View full abstract

Mechanical energy delivery during low-cycle system oscillation was studied by measuring the fluctuating pressure and flow rates at the inlet and outlet of a pump. By studying the relationship between the mechanical energy delivery versus flow patterns and observing the inlet and the inside of the impeller subject to cavitation, it was found that one of the causes of low-cycle system oscillation is the effect of cross sectional area reduction upstream of the impeller by the cavitation reverse flow. This is because a dynamic difference in phase between the cavitation volume versus inlet flow rate of the pump is a consequence of the above effect.

View full abstract

This paper presents the double linearization theory applied to lightly loaded two-dimensional subsonic and supersonic cascades undergoing oscillation with chordwise displacement or nonrigid deformation. Numerical computation based on this theory has been conducted to study parametric dependence of unsteady aerodynamic work on blades. Chordwise displacement can be favorable or unfavorable for stabilizing the translational oscillation, depending upon the phase difference between the chordwise and normal components of the blade motion. For supersonic cascades, the effect of displacement of Mach line reflection points due to the chordwise blade motion gives significant contribution to unsteady aerodynamic forces. The unsteady aerodynamic work for nonrigid blade motion is also substantially influenced by steady loading.

View full abstract

A new mechanism is constructed to transport fluids by means of a progressive wavy motion of two elastic plane membranes which face each other and cover two magnetic fluid layers adjacent to both side walls in a long duct of rectangular cross section. A steady fluid flow can be induced in an inner stratum which is formed between two opposite membranes. Wavy motions of membranes (the peristaltic motion) are excited by the external magnetic field in the form of sinusoidal waves propagated along a train of electromagnets which is arranged linearly along the axis of the duct and driven by a digital electronic pulse-width-modulation circuit and a DC power supply. As a result of examinations of fundamental characteristics of intensity of the propagating magnetic field, wavy motions generated on the membranes, and flow velocity and pressure induced dynamically, the mechanism proposed here exhibits very interesting the excellent performance to be useful in the practical electromagnetopropulsion and/or fluid transportation systems.

View full abstract

In this paper, a new theory for micropolar electrically conducting fluids is presented by unifying the theories of micropolar fluids and electrically conducting fluids. This theory makes it possible to treat both magnetic and dielectric properties simultaneously. It is expected to contribute to the knowledge of the properties of ionized gases, electrically conducting viscous fluids, including fine particles, and magnetic fluids. Futhermore, we give an exaxt representation of Maxwell's stress tensor in materials when volume strain arises from electromagnetic force. As a result, interaction between fluids and electromagnetic fields becomes cleal analytically. Thermodynamical quantities such as entropy, pressure, internal energy and enthalpy are shown as sums of the state function in the non-electromagnetic field and the additional state function due to the existence of the electromagnetic field.

View full abstract

Elasticities were measured under the conditions of shear and elongational flows of PEO and Separan solutions below the concentration level of several tens of ppm. The first normal stress difference in a shear flow was measured by utilizing reactions exerted by a jet issuing from a capillary, and the elongational stresses were also obtained separately with reactions or pressure drops produced in the flow through small orifices. The following points were clarified: the first normal stress difference τ₁₁-τ₂₂ had a relationship of the power law type with the shear rate γ, that is, τ₁₁-τ₂₂&prop;γ. For the PEO 5 ppm solution, the power law index s was 2.0, as predicted with the Maxwell model, and in this case, the relaxation time was determined as 0.0016 second. For the other solutions used, we had 1.2< s < 2.0 and the measured first normal stress differences could be expressed with a modified Maxwell model. The elongational stresses measured with the two methods of jet reaction and pressure drop generally agree with each other. The elongational stress can be expressed, albeit approximately, with the Maxwell model or the modified Maxwell model depending on the concentration of the solution used. In some cases, the elongational stresses of PEO solutions, at least qualitatively, agree with predicted values of the yo-yo model proposed by Ryskin.

View full abstract

A three-dimensional analysis method for a flow with liquid surfaces using the boundary element method (BEM) is developed. When the flow is treated as a potential flow and Bernoulli's equation is applicable the BEM is thought to be suitable for the analysis of flow with liquid surfaces. The method can be applied to sloshing problems in containers with liquid surfaces, and the collision of the liquid surfaces with the container ceilings can also be analyzed.

View full abstract

The starting jet flow from a large chamber with a constant pressure is simulated by solving full Navier-Stokes equations with primitive variables. The growth and development processes of a vortex roll up are visualized using marker particles and related to velocity and pressure fields which are difficult to obtain experimentally. The main results can be summarized : (1) a short period after starting, the isobaric pattern is a concentric circle centered at the exit slit, and the fluid issues radially from the slit ; (2) as time elapses, the separated layer initiates rolling up to a vortex with a local low-pressure area ; (3) the developing vortex is transferred downstream at a longitudinal velocity along the separated layer ; (4) at a certain lateral section, the situation is unsteady as the vortex approaches and is carried downstream. When the vortex is carried far downstream, the flow tends to steady ; (5) in the vortex area, the isobaric pattern is concentric where the fluid particles are vortically entrained.

View full abstract

Damage due to the coupled action of cavitation erosion and silt abrasion was experimentally studied using submerged jets in which artificial silts (Al₂O₃) were included. It was shown that the magnitude of the acoustic emission energy, impact pressure and material damage rate caused by the cavitating jet impingement were greater than those caused by the noncavitating one. Good correlations between the above three erosion intensity parameters were also confirmed. SEM photographs of damaged surfaces showed that the edge of the abrasion trace was partially broken into pieces by cavitation. The material surface was smoother and the abrasion pattern was changed compared to that of pure abrasion.

View full abstract

The instability of a thin liquid jet in a stationary atmosphere was analyzed by solving numerically the Orr-Sommerfeld equation for both sinuous disturbances and dilational disturbances. It was found that the thin liquid jet had hard mode and soft mode instability for both disturbances. Comparing theoretical results with experimental results, we discuss the possibility of the occurrence of the interfacial disturbances wave on the plane liquid jet. It could be suggested that the interfacial disturbance wave was mainly caused by the internal instability in the liquid jet flow and that the disturbance wave corresponded to the hard mode instability for the sinuous disturbance.

View full abstract

Approximate solutions to determine the two-dimensional subsonic flow through slits and 45-degree converging conical nozzles are described in this paper by expanding the subsonic Hodograph method into the finite opening area. The free stream line and contraction coefficient of jets through such constrictions are obtained. According to the comparisons between numerical exact solutions and present approximate solution of Hodograph equation, it is found that the present method is acceptable in practical use for evaluating free stream line and contraction coefficient except for in high-Mach-number flow regimes.

View full abstract

A fine striped interfacial wave pattern was observed on the thin liquid sheet jet in a cocurrent gas stream. The experiment revealed that this wave pattern affected the disintegration of a liquid sheet. In order to clarify the significant factor of the liquid jet wave motion, the fine interfacial waves on a plane liquid jet in a stationary atmosphere were also investigated experimentally. The interfacial waves were classified into the following five patterns: smooth, two-dimensional wave, modified two-dimensional wave, pebble wave and sandpaperlike wave. The fundamental mechanism of the fine interfacial wave was investigated by developing an optical measuring device.

View full abstract

Concerning with flow control and measurement of extremely small flow rates of a liquid, an experimental study on the mass of a falling droplet, m, from a capillary tube was carried out. The liquid used was water and methanol. The capillary tubes with outer diameters, D, from 1 to 10 mm, were made of teflon, brass and stainless steel. The main results obtained are as follows : (1) the mass of the droplet is given the following general equation regardless of the surface tension, σ, of the liquid, m/ρD³=2.15(σ/ρgD²)^1.06, (2) the above equation enables one to measure the flow rate for very small quantities.

View full abstract

The authors propose a method of measuring a 3-D flow field by digital image processing for pathline pictures. In this method, a 3-D flow field is measured by the measurements of two 2-D flow fields perpendicularly intersecting each other. A still camera, a color reversal film, halogen lamps, two flash units and red cellophane papers are used to photograph the pathlines of tracer particles in a measurement section. Each of the photographed pathlines in the measurement section has a brilliant white initial point, a bright red terminal point and a red middle part with intermediate intensity. The pathline picture on a film is twice entered into a personal computer by a video camera through a video memory. First, the initial point of the pathline and next the whole of the pathline are entered. The pathline picture is analyzed by the authors' original program and velocity vectors are obtained. As an application of this method, 3-D flow field in an agitated vessel was measured.

View full abstract

A numerical method is developed to solve three-dimensional incompressible viscous flow in complicated geometry using curvilinear coordinate transformation and domain decomposition technique. In this approach, a complicated flow domain is decomposed into several subdomains, each of which has an overlapping region with neighboring subdmains. Curvilinear coordinates are numerically generated in each subdomain using the boundary-fitted coordinate transformation technique. The modified SMAC scheme is developed to solve Navier-Stokes equations in which the convective terms are discretized by the QUICK method. A fully vectorized computer program is developed on the basis of the proposed method. The program is applied to flow analysis in a semicircular curved, 90°elbow and T-shape branched pipes. Computational time with the vector processor of the HITAC S-810/20 supercomputer system, is reduced to 1/10∼1/20 of that with a scalar processor.

View full abstract

The relaxation process to the equilibrium state for the d-type rough-wall boundary layer disturbed by a circular rod was investigated experimentally. The cylinder was positioned at two distances of y_c/δ₀=0.14 (inner layer) and 0.92 (outer layer). From the results, the mean velocity profile disturbed by the cylinder virtually recovered over a shorter distance for y_c/δ₀= 0.14 than that for y_c/δ₀=0.92. At the measuring station farthest downstream (x/δ₀=82.6), the mean velocity profile differed significantly from that of the undisturbed flow for y_c/δ₀=0.92. However, for both cases, the logarithmic law with x=0.41 was observed in the wide-region-of-flow field. The streamwise changes in the boundary layer characteristics, such as the skin friction coefficient, the roughness function and the wake strength determined from the measurements, indicate the different tendencies of the two cases.

View full abstract

A full three-dimensional inverse solution of the impeller blade shape is presented. The inverse method automatically calculates the blade shape which satisfies flow characteristics specified by a designer without trial iteration of flow analysis calculations. The flow is assumed to be inviscid and incompressible. The impeller blades are represented by superposition of vortices and sources on the blade mean surface. By expanding generalized velocity potential in a Fourier series, a set of two-dimensional quasi-Poisson equations in r and z is obtained. These equations are solved with SOR. Numerical examples are shown for the design of axial flow rotors with free vortex swirl. A blade thickness distribution and a pressure difference distribution across the blade are specified, and the blade mean-surface shape is determined by the inverse method. The results are compared with those presented in our previous report using a Fourier-Bessel expansion. The flow analysis for the designed impeller is performed by means of the two-dimensional singularity method and the three-dimensional panel method, and the calculation accuracy of the inverse method is examined.

View full abstract

In order to design a variable pitch angle windmill, which generates maximum torque at each angle, it is necessary to know the dynamic characteristics of an aerofoil. The dynamic characteristics of an aerofoil, whose pitch angle changes with constant speed in uniform air flow, are investigated experimentally. The characteristics are uniquely determined by nondimensional pitch angle speed cd/U, where c, a, U are chord length, pitch angle speed and velocity of flow, respectively. The maximum lift coefficient increases with ca/U almost linearly, and the shapes of hysteresis loops in the lift and drag coefficients are determined by ca/U.

View full abstract

The purpose of this paper is, first, to analyze 3-D transonic cascade flow by using the TVD Navier-Stokes code developed by the authors. Since the momentum equations of contravariant velocities are used in the code, the treatments of boundary conditions, especially the solid wall and the periodic boundaries for the 3-D cascade flow, become very easy. A Chakravarthy-Osher type TVD scheme modified by the authors is adopted in order to clearly simulate 3-D complex flow having shock waves and separation bubbles. The second purpose is to visualize the computed results by using the 3-D color graphic display. As numerical examples, the 3-D viscous flows thorough a transonic axial-flow compressor rotor with or without a tip clearance are computed, and the colored oli-flow patterns, 3-D particle paths, 3-D time lines which show the shock/boundary layer interactions and a leakage vortex near the casing wall are obtained. In particular, the mechanism of occurrence of the leakage vortex is investigated in detail.

View full abstract

The accurate prediction of rotor thrust is important when analyzing the strength of the rotor blades and supporting tower of the horizontal-axis wind turbines. In the present work, a series of model wind turbines were tested in a wind tunnel to ascertain the thrust of the rotor. The effect of three design parameters on the thrust ; i.e the number of blades, the pitch angle, and the up-tilt angle of the rotor plane, were tested for several wind speeds. The result of the experiments show that, on the condition of keeping two of the three design parameters unchanged, the thrust on the rotor decreases with increasing the pitch angle and up-tilt angle, and increases with increasing the number of blades. The results uncovered by the experimental analysis correlate well with the theoretical values. Moreover, the relationship between the thrust coefficient and the torque coefficient were also clarified.

View full abstract

An experiment was performed to investigate heat transfer around the second rectangular cylinder of three cylinders arranged in a line on a laminar boundary layer. The length of the cylinder was 30 mm, and the height was varied at 3, 5 and 10 mm and their pitch varied from 40 to 60 mm. The main results obtained were as follows: the heat transfer coefficient around the cylinder increased drastically as increasing in the free stream velocity. The local heat transfer coefficients decreased in the order of the front, upper and rear face. For laminar flow type, the dependency of the Reynolds number on the heat transfer was Nu &prop; Re^0.5, while for turbulent flow type, it was different in each face. There existed a correlation between the maximum turbulent intensity on the cylinder and the heat transfer coefficient.

View full abstract

A one-dimensional numerical model which simulates the dynamic thermal performance of stratified salt-gradient solar pond is described. The model simulates the thermal performance in the upper convective layer, the nonconvective layer and the lower convective layer in a cold climate. In addition to the energy flux in the solar pond, the model simulates the varying ice thickness appearing in the upper convective layer during the winter season. From the results, it is noticed that it is important to maintain the optimum thickness of the nonconvective layer and the clearness of the upper convective layer for the high thermal efficiency of the salt-gradient solar pond.

View full abstract

Large-scale turbulent structures in backward step flow are investigated from measurements of the cross-correlation between velocity fluctuation in the 3-dimensional flow field and pressure fluctuation on a reattachment plane. Large eddy structures and traces affecting a wall pressure are shown by sequential contours of cross-correlation as a function of time. The pressure fluctuation in a recirculation zone is mainly caused by the velocity fluctuations of a main flow, while in the reattachment region a pressure fluctuation is brought about by the velocity fluctuation entrained with a jet. This highly correlated region after the reattachment moves downstream with the same speed as the wall pressure fluctuation patterns visualized by real-time holographic interferometry.

View full abstract

In order to elucidate the effects of positive pressure gradient on flame structure and stabilization mechanism, it is first necessary to clarify those on turbulence characteristics and eddy structure not only in a cold shear layer, but also in a cold wake flow behind a bluff body. In this study, the influences of pressure gradient in the flow direction on the surface pressure around a circular cylinder and the eddy-shedding process are investigated in a two-dimensional cold air flow without combustion. Flow visualization and an FFT analysis of the oscillating surface pressure are made to examine time-averaged and fluctuating properties concerning the eddy-shedding process. It is found that increasing the free stream pressure gradient causes a transition of the wake flow pattern from an alternative eddy-shedding type to a symmetric one without Karman vortex. An increase in the pressure gradient extends the wake region, and simultaneously lowers the pressure drag force, which may be regarded as one of the most important characteristics of the divergent flow.

View full abstract

OH fluorescence induced by a XeCl excimer laser is detected in a laminar diffusion flame. OH concentration and temperature in the flame are estimated by OH fluorescence and Rayleigh scattering. The results obtained are as follows. (1) The spectrum of OH fluorescence is numerically simulated by rate equations and it predicts the experimental results well. (2) Relative and absolute OH concentrations are estimated by the measured fluorescence and rate equations. (3) Rayleigh scattering is separated from OH fluorescence to measure flame temperature. The measured temperature agrees with that obtained by use of the Rayleigh scattering by Argon ion laser.

View full abstract

In the evaluation of LDA signal processing, numerical simulation experiments were carried out. The burst digital correlation (BDC) and the burst spectrum analysis (BSA) were examined and compared with each other. The simulation results show that the BDC can process Doppler signals of SNR above -12 dB while a limit of -10 dB was found for the BSA. A bandpass filter was used. The use of the filter extends the SNR limit of the the BDC by 3 dB. A coincidence circuit can improve the performance of the BDC in the correlation procedure. The present computer simulation is found to be valuable for the evaluation of LDA processors and the prediction of their performance.

View full abstract

The thickness of the burning zone and the turbulent burning velocity in the cylinder of the sparkignition engine were compared with those in constant volume vessels of the disk and cubic types. As a result, it was found that the thickness of the burning zone had a good correlation with the value of the relative burning velocity which was the turbulent burning velocity divided by the laminar burning velocity, under various conditions of fuel (propane and methane), equivalence ratio (0.8∼41.4), pressure (atmospheric pressure∼1.6 MPa) and unburned gas temperature (300∼620 K). Moreover, the turbulent flame in the burning zone could be considered to be the burning of islands based on the discussion of the relationship between the Reynolds number and the Damkohler number in this engine except for conditions of very weak turbulence.

View full abstract

The performance of a spark ignition engine operated with various fuels was calculatedly estimated by the computer model. The difference of thermal efficiency and NO emissions among fuels was derived mainly from the difference of flame temperatures. The cycles operated on alcohols or paraffines yeild high thermal efficiency and low emissions of NO owing to their comparatively low flame temperatures. On the contrary, when acetylene, hydrogen or benzene is used as a fuel, the flame temperatures rise high enough to cause the deterioration of the efficiency because of thermal dissociation, resulting in fairly high emissions of NO, especially with lean mixtures. Provided that hydrogen is used under its stoichiometric mixture, the NO emissions are relatively low.

View full abstract

An experimental study was made to determine the time history of fuel vapor concentration in the combustion chamber of an automotive spark ignition engine. The laser Rayleigh scattering was applied for the remote, nonintrusive, point-probing of the concentration of fuel vapor in the combustion chamber, which was caused by the timed injection of liquid fuel into an intake port. In order to make the optical diagnostics accessible, the original engine was modified to permit installation of an extended combustion cylinder with three sets of glass windows and an extended piston. The results showed that the vapor concentration increased and reached a peak after which it decreased during the intake stroke, and was almost constant during the compression stroke. It was also found that the vapor concentration increased with a reduced air fuel ratio supplied, and with increased engine speed and fuel injection pressure. A maximum occurred as a function of fuel injection timing.

View full abstract

A chemical process capable of reducing nitric oxide from the diesel engine exhaust is studied. In this process, an amidogen (NH₂) -containing ammonia gas is mixed with the exhaust gas to ensure destruction of NO by NH₂ at a relatively low temperature level. Theoretical predictions based on the rate-controlled partial equilibrium method show that the addition of only a small amount of NH₂ may effectively reduce nitric oxide in the exhaust gas. The feasibility of the proposed process is demonstrated on a miniature system consisting of an NH₂ producer and a mixing chamber into which a part of exhaust gas from a diesel engine is introduced. Results show that a significant reduction of nitric oxides is achieved at a mixing-chamber temperature higher than 750 K. Furthermore, the practical feasibility in controlling NO_x is compared with that of the cianuric-acid method proposed by Perry and Siebers.

View full abstract

Measurements of cylinder bore radial deformation were made along the peripheral direction during actual engine operation. A piston, whose upper part was capable of turning, was installed on the pin boss and was slowly turned from an outside power source by means of a gear train. A heatproof inductance-type gap sensor was embedded on the second land surface. When the engine was under steady-state operation, the piston position at any crank angle was fixed every cycle, not only axially but also inside the corresponding cylinder cross section. As a result, the cylinder bore deformation was continuously measured by one sensor along its entire periphery at one desired axial piston position. One of the important measurement results was that the compound effect of head bolt force and thermal stress had a significant influence on the bore deformation.

View full abstract

Power output performance of the rotary engine is improved when the dynamic effects of intake and exhaust systems are utilized. Although one-dimensional gas exchange process simulation is effective to efficiently develop the systems utilizing the effects, the simulation was difficult to apply to the peripheral port-type rotary engine because of the simultaneous gas exchange between one port and two combustion chambers. A boundary model connecting the pipe and the volumes were developed. This report show the validity of the model, and successful applications of the simulation to the engine.

View full abstract

In order to improve the accuracy of the estimation of swirl intensity, the treatment of the valve outlet boundary has been improved. In the improved treatment in this study, the gas flow is calculated by taking into account the configuration of the valve face and the valve seat using the porosity approach, without prescribing the outflow direction. By comparing the calculated results with the experimental results under unsteady conditions, it has been found that this improved treatment enables one to estimate the dynamic effects of the intake pipe on swirl intensity and volumetric efficiency. By using this improved treatment, numerical prediction has been carried out under motoring conditions of a practical engine. The calculated results reveal that the swirl ratio varies considerably with the engine speed and that the variation of the swirl ratio does not correspond to the variation of volumetric efficiency with the engine speed.

View full abstract

Numerical simulation was carried out to clarify three-dimensional air motion in the swirl chamber of a diesel engine. Simulations have been performed spherical swirl chambers with different sorts of connecting ports. The results obtained are as follows. (1)Toward the end of the compression stroke, the swirling flow becomes strong. On the other hand, the flow normal to the swirling surface becomes comparatively weak. (2)According to the jet of a subconnecting port, the swirling flow in the end stage of the compression stroke becomes weak in the central part. (3)The spatial distribution of strong turbulent intensities corresponds to the strong flow field and the shear force field based on inlet flow.

View full abstract

The droplet size of diesel fuel spray has been studied experimentally using a marine dieselfuel-injection valve, a constant pressure fuel-injection system, a high-pressure vessel, and a particle size distribution analyzer. Measurements were carried out at many points in the fuel spray for wide variations of fuel-injection pressure, up to 130 MPa, and ambient air pressure, up to 2 MPa. It has been found that the droplet size distribution of the fuel spray is very narrow and does not vary with radial measuring points in the fuel spray when the fuel is injected in the atmosphere with high injection pressure. The spray is truly mono-dispersed. As the ambient air pressure is increased, the mean droplet diameter increases and the droplet size distribution becomes wider.

View full abstract

It is unavoidable that a DI diesel engine exhausts a blue and white smoke at starting, especially in the cold atmosphere. In the report presented here, a small DI diesel engine was started under the conditions of the coolant and suction air whose minimum temperatures were -18°C and -5°C, respectively. The flame was photographed by high-speed photography, the temperature of flame and the soot concentration were measured by two-color method, and CO₂ concentration was detected by the luminous method. The engine cannot be started over several cycles when the coolant temperatures of coolant and suction air are decreasing, the maxima of the cylinder pressure, flame temperature, soot concentration and CO₂ concentration are decreasing. Luminous small dots or a small lump of flame become scattered in the piston cavity.

View full abstract