In the incompressible viscous fluid flow analyses, the backward-facing step flow problem has been investigated as a benchmark problem with the inflow and outflow boundary conditions for testing the effectiveness of a new numerical scheme. In this research, the corrected fourth-order finite difference method is employed for this fluid flow analysis. This method was originally proposed by authors through the error analysis approach and has been successfully applied for viscous fluid flow and natural convection problems in a square cavity. In this study, the backward-facing step now was simulated for the Reynolds numbers of 800 and 1000, and the steady solutions were obtained. The effects of the sizes of computational meshes to the reattachment lengths are investigated, and the mesh-independent solutions are obtained. We confirm the agreement of our results with other research works and the effectiveness of our corrected fourth-order finite difference method for viscous fluid flow analyses with inflow and outflow boundary conditions.

抄録全体を表示

The flow field in the turbo-molecular pump is numerically analyzed by the Direct-Simulation-Monte-Carlo method. The two dimensional simulation shows these results : 1) Even at the steady state of the pump, some unsteady now exists at the area between the rotor and stator. 2) In the pump with a rotor stage and a stator stage, the relative velocity is not similar because of the boundary condition of discharge side. 3) The maximum pressure ratio is effected by the time-averaged flow in the pump. 4) Intermolecular collision makes the maximum pressure ratio small at Kn < 10, but the discharge time is not affected by the molecular collision. 5) In the pump with many stages, the flow becomes almost similar when the shape of blades are same.

抄録全体を表示

The noise prediction of scroll compressor is one of the most important subjects to develop a more calm and efficient air conditioner. A scroll compressor has a finite volume ratio, which Is defined by the scroll shape. Because of this finite volume ratio, the noise increases in the high-pressure ratio conditions. This study is the first step to predict the noise, which is caused by pressure pulsation. In previous works, we analyzed the discharge pressure pulsation by CFD with 1-dimensional straight tube model, which had same cross-sections. However the model was in need of the experimental values such as the equivalent sound length of compression chambers. From this reason it was impossible to predict the frequency pattern of noise. In this stydy, we developed a new 1-dimensional tube model, which has variable cross-sections to consider the change of compression chamber shapes. From analysis of the pressure pulsation frequency spectrum, it is confirmed the new model can predict reasonably the change of the sound power level by the setting of the lead valve.

抄録全体を表示

One-dimensional flow model for non-Newtonian fluids in a single cavity die with a slot is presented. The equations of mass and momentum in a cavity are treated one-demensionally by assuming an appropriate mean flow over the cross section of the flow. The flow field inside the slot is assumed to be fully developed. The equations of flow for the cavity and the slot are derived and then coupled. We use a finite difference method to solve these governing equations. It is concluded that cavity's geometry and fluid properties affect the uniformity of the flow from the slot, and the presented model agrees with the experimental result.

抄録全体を表示

The blood flow in large arteries is commonly analyzed by means of the constitutive equations. However, it is not appropriate to use constitutive equations for small arteries because of the heterogeneity of the blood. In this paper, a new method to model an erythrocyte by using beads, springs and dampers is proposed as an alternative to analyze the blood flow. The behavior of a single erythrocyte is computed under an oscillatory shear field. The real and imaginary part of dynamic viscosity, dynamic modulus and normal stress difference are discussed. The results show that the bead-spring-damper model appropriately can express the dynamic characteristics of blood. It is therefore considered that the bead-spring-damper erythrocyte model is able to consistently express blood flow without using constitutive equations.

抄録全体を表示

The characteristics of unsteady flow generated by sinusoidal motion of fluid in a circular pipe connected perpendicularly with the wall of a large vessel have been investigated experimentally. The time-dependent velocity fields near the opening of the pipe are obtained systematically by the visualization method for a wide range of the Womersley number a and the Reynolds number Re, and also measured minutely by a laser Doppler velocimeter for flows at Re=1 000 and a=10∼25. The results show that the flow Patterns are classified Into three types judged from the behavior of fluid particles passing through the pipe opening. For low Womersley numbers of a^2.1/Re < 0.25, the flow out of the pipe produces vortex rings in the vessel and fluid different from that discharged into the vessel comes into the pipe along the vessel wall. For sufficiently high Womersley numbers of a^1.8/Re > 0.29, on the other hand, the flow consists of a circulation around the pipe edge and axially reciprocating motion through the central part of the pipe opening. An explanation is provided for the formation and the flow behavior of the vortex rings, the features of which are dependent on a and Re.

抄録全体を表示

Flow generators which take advantages of a flexible flat plate with its leading edge shaken perpendicularly in a uniform now, which have recently been interested in and/or been begun to be utilized in several purposes, for example, flow thrusters, now mixers, and fans. The behaviors of the plate and its characteristics were analyzed on the assumptions of potential now and infinitesimally small perturbed motion. The characteristics of the fluid dynamic performances and behaviors, including propulsion efficiency and thrust coefficient, have been given in terms of the mass ratio between the plate and the fluid, the plate stiffiness/flow force, reduced frequency, relative frequency, etc.

抄録全体を表示

Overall performances and fluid-elastic characteristics of Vibrating Elastic Plate Flow Generators were analyzed on the method described in the first report⁽¹⁾ of this investigation. The performance characteristics, including propulsion efficiency and thrust coefficient, have been given in terms of mass ratio between the plate and the fluid, plate stiffness/flow force, reduced frequency, relative frequency, etc. From the results, a map for optimization of the flow generators was constructed. The optimum conditions were found to be achieved by merging of both a fluid-dynamically favorable matching between the plate orientation and the flow and a favorable elastic deformation of the plate.

抄録全体を表示

Using a quiet jet facility the relationship between self-excited oscillation of the round jet impinging upon a pipe's leading edge and its radiation field was studied experimentally in Reynolds number based on momentum thickness range 144≤Re_θ≤502. A pipe's leading edge intercepts a laminar free shear, the shear layer can be into oscillation, depending on the convectional speed and the regular interval of discreat vorticies. The oscillation can produced resonance of air column in a pipe as the oscillation mode become characteristic frequency of pipe length with increase impingiment length. The principal results are as follows : self-excited oscillation were observed in non-dimensionalized dominant components and impingiment length range 0.0028≤βθ/u≤0.21, 18≤l/θ≤380, they are associated with the nonlinear saturation of the unstable wave amplitudes of the shear layer occuring at the vortex-vortex interaction ; the locations and acousitic power of the resonance can estimated from emprilcal curve in figure 4 and emprical equation (2) ; resonance tone are completely disappeared when the free shear layer become to turbulent by enhancement of the tripping wires.

抄録全体を表示

An experimental investigation attempting to actively control the flow over a cavity and to reduce the aerodynamic noise is carried out. Thin piezoelectric pieces are used as the actuators. Two types of piezo-ceramic devices, a uni-morph type and a bi-morph type, are compared. Experimental results show that both types are capable of generating a flow field that is 180 degree out of phase along the spanwise direction. However, the uni-morph type piezo-ceramic actuator pieces can control the flow under wider range of experimental conditions. The unimorph type actuator pieces are also more effective in suppressing the aerodynamic noise.

抄録全体を表示

Surface pressure fluctuations before the onset of separation were investigated experimentally in order to find a precursor signal of separation. To generate managed separation, an inclined wall of a diffuser was opened with time at a constant angular velocity. Separation occurred when the angle of the inclined wall reached at 13 degrees in our experimental condition. Sinusoidal wave amplified with time before separation was detected by a pressure transducer mounted on the Inclined wall. The frequency of the wave was obtained by FFT analysis applied to small intervals in time of the pressure fluctuations. The particular frequency was found to be related to the instability wave due to an inflection point of velocity profile. This type of velocity profile appeared just before separation. Therefore, the sinusoidal wave can be regarded as a precursor of separation.

抄録全体を表示

The radial motion of a spherical vapor/gas bubble in acoustic fields is investigated numerically. A mathematical model is derived with an assumption of the spatial uniformity of pressure in the bubble. The model includes the effect of spatial distribution of temperature and concentration, and heat and mass transfer at the bubble wall. The small-amplitude bubble oscillation in a hydrothermal system is simulated. The numerical results show that the spatial uniformity of pressure is valid. The amplitude of radial motion is correlated well with the nondimensional interdiffusion coefficient D_gv/ωR₀² where D_gv denotes the interdiffusion coeficient, ω the driving frequency, and R₀ the initial bubble radius. The phase change and gas diffusion at the bubble wall interact to each other. The interaction enhances the mass transfer at the bubble wall.

抄録全体を表示

A new smoke-wire technique was examined to get quansitative information of turbulent flow fields from visualization photographs. Smoke emission on the smoke-wire was regulated by partially plating to give it locally different Joule heating. The smoke streaks emitted from unplated points of the wire captured the flow characteristics quite clcarly. When the small delay time between the smoke emission and the exposure was given, smoke streaks directly indicated velocity vectors at each point in the flow field. A smoke-wire grid composed of 20 partially plated smoke-wires was installed behind a circular cylineder and a delta-wing. Conditional visualization was conducted to clarify the behavior of coherent vortices behind the circular cylinder based on a velocity signal from a reference hot-wire probe filtered at the vortex shedding frequency. Phase-averaged velocity vectors and vorticity fields were calculated from the conditioned smoke streaks in the wake of the cylinder and the trajectories of the coherent vortices were determined.

抄録全体を表示

The direct numerical simulation (DNS) offers useful information about the understanding and modeling of turbulent flow. Few direct numerical simulations of wall-bounded compressible turbulent flows have been performed however. The objective of our study is to construct the DNS algorithm which can simulate the compressible turbulent flow between adiabatic and isothermal walls. The flow is selected since it is the simplest turbulent flow with an adiabatic wall, and therefore it is ideal for the modeling of compressible turbulent flow near the adiabatic wall. From the applicability and efficiency, we construct the new DNS algorithm based on the B-spline collocation method in the wall-normal direction. The validity and efficiency of the DNS algorithm are confirmed by comparing our results with those of an existing DNS in the compressible turbulent flow between isothermal walls. Then the DNS of compressible turbulent flow between the adiabatic and isothermal walls exhibits the applicability and usefulness of the algorithm.

抄録全体を表示

A pair of turbulent spots was simultaneously generated in a zero pressure-gradient laminar boundary layer on a flat plate. Multi-hotwire measurements were conducted to obtain instantaneous configurations of the turbulent spots throughout their merging process. The spot is composed of a number of hairpin-like vortices. At each wingtip of the spot, an isolated longitudinal vortex induced an upwash which conformed a velocity defect region above it. As the two simultaneously generated spots grew downstream, they began to merge at their inside wingtips. Mutual interaction between the longitudinal vortices at the Inside wingtips enhanced the velocity defect above the merged region, where the resultant unstable velocity inflection gave birth to stronger spanwise vortices with the longitudinal vortices beneath. These spanwise vortices developed into large-scale horseshoe vortices in the transitional boundary layer.

抄録全体を表示

As fundamental experiments of a sand collector with an injection port, the effect of the dimension of the sub port (the length of the sub hole and the radius of the main hole) in establishing the sub port in the injection port was examined on the collecting performance and flow visualization. As a result, the improvement on the collecting volume concentration was obtained, when the sub port was established In the injection port further than the conventional case. The concentration rises, as the length of the sub hole is shorter, and as the radius of the main hole is smaller. From the visualized image of the fluidized region of collected particles, the fluidization of the jet blown out from the sub port was able to be observed, and it was possible to prove the relation between the above-mentioned concentration.

抄録全体を表示

Experimental studies are made on the convective heat transfer of three types of thin porous bodies. Heat transfer performances, flow patterns and temperature profiles near the porous bodies are compared with each other. The heat transfer performance of porous bodies with the largest pore diameter is large. It became clear that the high heat transfer performance depends on an excellent heat transportation ability inside the pore and near the surface of the porous bodies.

抄録全体を表示

If a flow obstruction such as a spacer is set in a boiling two-phase flow within an annular channel, the inner tube of which is used as a heater, the temperature on the surface of the heater tube is severely affected by the existence of the spacer. In some cases the spacer has a cooling effect, and in the other case it causes the dryout of the cooling liquid film on the heating surface resulting in the burnout of the tube. But the thermo-fluid dynamics mechanism to cause burnout near the spacer is not still clear. In the present paper we discuss temperature fluctuation characteristics in relation to the change of the flow configuration near the spacer in transient boiling two-phase flow. The transient parameters are the heat flux, the mass flow rate, and the inlet quality.

抄録全体を表示

The learning process of neural network in the impedance CT for void distribution measurement was improved. A void distribution function to express void distribution only by three parameters was introduced, so that 81 of output layer units in the previous network, was decreased to 3 units. Input layer units were reduced by eliminating duplicated current data and numbers of hidden layer units were minimized. The best learning parameters were recalculated about the slimed network by using the error back propagation technique. As the results, the learning speed of the impedance CT was about 30 times faster than that of the previous paper. Two types of experiments were conducted using a square duct. One used acrylic rods as dummy voids, the other used air-water two phase flow. Comparison between experiment and prediction indicated that the impedance CT method was able to predict void distributions with good accuracy.

抄録全体を表示

Point of net vapor generation (PNVG) was investigated based on the void fraction dataset obtained by using high-frame-rate neutron radiography. The test channels used in the experiment were rectangular channels heated from one side with gap thickness of 3 and 5 mm, channel width of 130 mm, and heated length of 100 mm. This study focused on (1) the instantaneous and time-averaged PNVG, (2) the effects of system parameters on PNVG, (3) the applicability of existing PNVG correlations to the channels with short heated length, and (4) effect of the PNVG in CHF, model. The following results were obtained : (a) the effects of system parameters on the thermal, equilibrium quality at the PNVG are not significant in the present conditions, (b) existing PNVG correlations tend to underestimate the thermal equilibrium quality at the PNVG, and (c) the accuracy of Katto's CHF model could be improved significantly by calculating the PNVG with high accuracy.

抄録全体を表示

The high heat flux in MEB far exceeding CHF should be considered a result of violent growing and collapsing behavior of coalescent bubbles in that region, which strongly introduces subcooled liquid to the heated surface. This bubble behavior generates a high pressure in the vicinity of the surface. In this study, a measurement of the pressure fluctuation in the liquid near the surface was performed synchronously with a recording of the bubble motion by use of a void probe set close to the surface. In high liquid subcooling conditions (S-MEB I), the void probe signal in the vicinity of the surface shows regular cyclic behavior synchronous with the pressure waveform. Under a medium subcooling condition (S-MEB II), there appear three different modes in the void signal in sequence. The one looks random and does not synchronize with the pressure waveform. The another is very systematic and similar to the signal in high liquid subcooling. The other shows a constant void fraction of a high value, indicating the surface to be covered with a thick vapor blanket. The time fraction of the second mode increases gradually with increasing the heat flux.

抄録全体を表示

A wake from an elliptic cylinder with fixed separation points is analyzed numerically, and is clarified as follows : An isothermal Karman vortex street never occurs at Re<36, but a cooled Karman vortex street is generated by cooling a progressive wavy wake at 26≤ Re<36, where Re is the Reynolds number. The cooled Karman vortex street can be generated at smaller value of the absolute Richardson number in an elliptic cylinder wake than a circular cylinder wake. By cooling the Karman vortex street, a suppressed flow Is generated. Both the Karman vortex street and the suppressed flow have a double structure combined the motions due to both negative buoyancy and the forced flow. The dominant motion is due to the forced now at the Karman vortex street, and due to negative buoyancy for the suppressed flow. At the boundary between the Karman vortex street and the suppressed now, the dominant motion is exchanged, and the wake frequency and the mean Nusselt number decrease abruptly.

抄録全体を表示

This paper describes the velocity and acceleration of a liquid drop exposed to an airstream. Experiments were conducted using a 60 mm×150 mm cross-sectional and 500-mm-long horizontal air-suction wind tunnel. Water was fed to a nozzle attached to a vibrator. Naturally dripping uniformly sized drops from the nozzle or uniformly sized liquid drops produced by a longitudinal vibration of the nozzle were used. Drops were exposed to airstreams and moved downstream. The prediction of the trajectory of droplets is useful for designing high-speed aircraft, combustion facilities, cooling facilities, etc. In order to predict the motion, the drag coefficient is required. However, the drag coefficient of a deforming liquid droplets has not been investigated sufficiently. To calculate the drag coefficient, the acceleration of a drop must be known. Therefore, in this experiment, the displacement and deformation of liquid drops exposed to airstreams were measured from photographs obtained by triple exposure photography or consecutive photography. From these data, velocity and acceleration of a drop exposed to airstreams were calculated.

抄録全体を表示

The blue-name reaction is a real issue for the autoignition of hydrocarbon fuel/air mixtures in internal combustion engine cylinders. Once a blue flame has appeared, the following hot-flame onset would be inevitable, except for very special cases. Blue flame generates carbon monoxide briskly, and is the final induction stage of oxidation up to the real hot-flame ignition. Piston-compression ignitions of n-butane/air mixtures conditioned at slightly upper and lower than the lean ignitable pressure limit were compared using a rapid compression machine to elucidate the ignition promoter or ignition trigger to the final hot-flame appearance. Carbon-monoxide concentration is always superior to of the carbon dioxide during the whole induction period and its maximum is caused at the very late stage of blue-flame period. Based on chemical species histories it could be concluded that a carbon-monoxide/dioxide ratio should increase for the hot-flame establishment. A small amount of carbon-monoxide brimming over compared with carbon dioxide production rate during the blue-flame period would be a trigger for the transfer to the final hot-flame ignition.

抄録全体を表示

A control method of crystal growth for diamond films has been established with improving the surface quality by use of highly reducing name with high-voltage pulse discharge. The nucleation density, the crystal diameter and the surface roughness can be controlled actively by applying the high-voltage pulse on the substrate. Especially, it is also possible that pulse current over 40 mA changes the crystal surface direction. When ion flux flows to substrate, the nucleation density of diamond increases with the pulse voltage, and then smoothed surface films can be obtained. On the other hand, when electron flux flows to substrate the surface morphology hardly changes. This is due to the CH₃⁺ ion flux, which is considered to be diamond precursor, increases with the electric field and become higher than CH₃ thermal flux multiplied the sticking probability. This phenomena results in the increase of supersaturation of the radicals on the substrate and in the enhancement of nucleation. Simultaneously, CH₃⁺ sticking coefficient itself increases, because of both high incidence speed of CH₃⁺ flux to the substrate and high directionality.

抄録全体を表示

In a discharge denitrification process where discharge electrons collide with combustion gas to produce N, O, OH and H radicals, the produced radical is said to be a key factor which decides the denitrification performance through the oxidization and/or the reduction of NO_x. In this study N, O, OH and H radical densities have been measured by the appearance mass spectrometry in a low-pressure discharge tube In which the simulated combustion gas flows. The radical densities of about 10¹¹-10¹² cm^-3 have been measured at the outside of cathode fall region when the s-n ratio is high. Boltzmann's equation for the electron energy distribution function has been solved numerically under conditions of the measured electric field and the experimental gas composition, and the denitrification chemical reactions have been calculated kinetically including the produced radical densities. The measured radical densities agree well with the analytical which correspond to the same electron density as the measured by the Langmuir probe.

抄録全体を表示

The flame reignition phenomenon that occurs after blowout of main flame is studied by using hydrogen fuel ejected vertically upward from a straight circular nozzle. Schlieren technique is employed to visualize the hydrogen flame, with focus on the region near the nozzle exit. It is found that : (1) The flame reignition phenomenon of hydrogen jet diffusion flames occurs only in certain nozzle diameters and flow rates ; (2) The small-sized flamelet (2∼3 mm) remains on the nozzle rim at the mass flow rates in which the reignition occurs. A further increase in mass flow rates causes the flamelet to extinguish and no reignition appear any more ; (3) The interval of name reignition increases with an increase in mass flow rates ; (4) Rim thickness does not affect the flow rates of reignition start and reignition finish. However, it influences the existence of flamelet formed above the nozzle rim.

抄録全体を表示

Upstream velocity field and curvature of a triple flame generated with methane-air mixture are experimentally investigated in various fuel concentration gradients. The flame curvature, which is measured from visible flame images, rises by increasing the concentration gradient. PIV data indicate that the triple flame burning velocity is almost constant. The flame strain rate goes up with the increase in concentration gradient. C. J. Sun et al. propose the theory on the burning velocity, which takes into account the effect of flame curvature, strain and Le number. The burning velocity by their theory is smaller than the measured triple flame burning velocity and shows the tendency to decrease by increasing the concentration gradient. This result indicates that there is another effect which strengthens the burning intensity. The effect also cancels the others, which weakens the burning intensity. There is tendency for this effect to go up by increasing the concentration gradient. It suggests presence of the flame interaction between the premixed and non-premixed flames at a leading edge of the triple flame.

抄録全体を表示

The purpose of this paper is to investigate the turbulent combustion properties of CH₄/O₂ mixtures diluted by CO₂ compared to those of CH₄/O₂/N₂ mixtures to study the feasibility of new combustion method, which is expected to be effective in the NO_x reduction and the improvement of combustion. Experiments are conducted by using spherical combustion bomb for applications to internal combustion engines. Consequently, it is shown that CH₄/O₂/CO₂ mixtures have larger turbulent burning velocity as compared to CH₄/O₂/N₂ mixtures under almost the same turbulent condition. Furthermore these mechanisms are discussed in view of change in local burning velocity due to flame stretch. As a result it is concluded that the differences in Markstein number between the CH₄/O₂/CO₂ mixtures and CH₄/O₂/N₂ mixtures, which are mainly due to the differences in Lewis number, cause the relative change in local burning velocity and this leads to the relative increase in the turbulent burning velocity in favor of CH₄/O₂/CO₂ mixtures.

抄録全体を表示

Systematic and logical similarity analyses of the flow, mixing and combustion were conducted based on the fundamental conservation equations of mass, momentum, species and energy, and boundary conditions that govern their phenomena. The dominating dimensionless parameters were derived for discussing the similarity law of the flow, mixing and combustion in turbulent flow field based on the assumption that the turbulence is strong enough and the influence of laminar transport can be entirely neglected. Therefore, the k-ε two equations turbulence model and the transport equations for k and ε were taken into consideration when deriving the dominating parameters. And, also the turbulent reaction model was incorporated when the dominating parameters for the turbulent jet flames were derived. Some examples are show of the similarity of non-reacting turbulent jets and turbulent jet flames. Their dominating parameters of the flame height of turbulent jet flames are also discussed.

抄録全体を表示

Combustion characteristics and engine performances of a swirl-chamber type IDI diesel engine using two-stage injection were experimentally investigated. The injection timing and fuel allocation ratio between first and second injections were widely varied. Heat release rate was analyzed to evaluate the combustion phenomena in the engine. As the result, ignition delay period for two-stage injection became longer than that for single-stage injection when the amount of fuel in the first stage injection was small. However, when the amount of fuel in the first injection was large, there was no difference of ignition delay period between two-stage injection and single-stage injection. NO and Smoke emissions and fuel consumption were improved by two-stage injection when ingition delay period became long.

抄録全体を表示

A three-dimensional simulation technique for stratified combustion process in direct injection gasoline engines is developed. The laminar flame speed for wide range of mixture equivalence ratio and EGR condition is modeled taking into account the reference temperature Intermediate between unburned and flame temperature for chemical reaction. This new laminar flame speed model and the coherent flame model are incorporated into a CFD code. The calculated flame propagation process, heat release rate and exhaust emissions are validated by measurements including LIF technique. The good agreement obtained for various operating conditions shows the availability of this method.

抄録全体を表示