The appcarance of a temporal irrcgularity of acrodynamic force of a rectangular cylinder with the critical depth (D/H=0.6), seems to be such startling phenomenon that is well worth investigating further. In this paper, turbulent flow around a stationary rectangular cylinder with the side ratio D/H of 0.2 to 1.0, has been computed by the LES method at high Reynolds number of 2.2×10⁴. The Smagorinsky model was used as a subgrid scale (SGS) model. The main objcctives of this work are to predict and evaluate characteristics of flows for a stationary cylinder. The computational results of force coefficients such as mean and fluctuating values of drag and lift forces, Strouhal number and flow-configurations correspond well with experimental data. We confirm that the abrupt changes of the aerodynamics forces appear irregularly with the time for the cylinder with a critical section. The temporal variation of a drag force between low and high values is certified to be accompanied with the corresponding change of instantenious frequency of lift force, by using of a wavelet transform method and an analysis of frequency histogram. The time history of drag force C_D agrees well with variation of the intensity of the wavelet transform coefficients or Strauhal number from the wavelet analysis results of lift force C_L.

抄録全体を表示

It is important to predict flow-induced vibration of a bluff body elastically supported in a uniform flow. In this paper, the free oscillation of a rectangular cylinder with two-degree of freedom in the streamwise (in-line) and cross-flow (transverse) dircctions in a uniform flow, has been computced by the LES method at high Reynolds number of 2.2×10⁴. The Smagorinsky model was used as a subgrid scale (SGS) model. The main objectives of this work are to predict and estimate characteristics of flows around a free-ocillating cylinder. The present computations successfully reproduce various types of flow-induced vibrations of a free-oscillating rectangular cylinder which were found by experiments; in-line oscillation, eddy-excitation and low-velocity galloping.

抄録全体を表示

Flow around a circular cylinder in oscillatory flow was studied by a numerical simulation using a finite-volume method. The computations were carried out by assuming 3-dimensional (3-d), unsteady, incompressible and viscous flow through Keulegan-Carpenter number (KC) in the range of 3≤KC≤25 at Stokes number (β)=95. The distribution of transverse force coefficient C_L corresponds well to the shape of spanwise cell structure formed on a cylinder surface at KC=3, 4. At KC=20, there appear many peaks of dominant frequencies in the power spectrum of in-line force coeffient C_F and there also appears the decrease of spanwise correlation coefficient of C_L, which cause to decrease the rms values of in-line and transverse force coefficients C_Frms and C_Lrms. The drag and inertia coefficients of Morison equation C_D and C_M and C_Lrms are compared with the measured ones and their results are in good agreement. It was found that C_D is the maximum value and C_M the minimum when twin-vortices are formed.

抄録全体を表示

Experimental and numerical studies on the vortex breakdown in a differentially-rotating cylindrical container are presented. Flow depends on the rotational Reynolds number Re_Δ=R²(Ω_t-Ω_sb)/v based on the relative angular velocity, the aspect ratio of the container H/R and the Rossby number Ro=|Ω_t-Ω_sb|/2Ω_sb or the Ekman number Ek=v/(2Ω_sb R²), here v is kinematic viscosity of working fluid, Ω_tand Ω_sb respectively angular velocities of top endwall and container (sidewall and bottom endwall), H fluid depth, and R radius of the endwalls. In despite of the rotation of the container was very small compared with that of the top endwall, flow was quite different from the case without rotation of the container. As Ek increased, the breakdown bubble moved toward downstream for the case of counter-rotation (Ω_tΩ_sb<0), whereas the breakdown bubble moved toward upstream for the case of co-rotation (Ω_t-Ω_sb>0). The boundaries of vortex breakdown for the cases of non-rotation, co-rotation and counter-rotation of the container were shown in the flow regime map. The necessary condition for the appearance of vortex breakdown which proposed by Brown and Lopez was applied to the present cases. Consequently, the influence of Coriolis force on the necessary condition was made clear for the cases of co-rotation and counter-rotation.

抄録全体を表示

Vortex shedding from a rectangular cylinder of which section has various width and height was studied numerically by using Large Eddy Simulation (LES) with Smagorinsky model. Simulations were performed for cases where width/height ratios are 1.0, 1.5, 2.0, 2.3, 2.5, 2.8, 3.0, 3.5 and 4.0. We focused on details of vortex shedding at the ratios of 2.3, 3.5 and 4.0 by visualization. As a conclusion, we considered the relations between Subgrid-Scale (SGS) dissipation and Grid-Scale (CS) enstrophy at each the ratio. They are correlated in various flow patterns for all the cases. Values of SGS dissipation in developing vortices are lower than values of SGS dissipation in the separated shear layers. On the other hand at a large ratio (3.5), the vortex coupling is occurred at the rear corner of the cylinder and the value of SGS dissipation of this case is 6.2 times higher than the value before coupling because of distortion of vortex.

抄録全体を表示

In the first report, the two-dimensional bubble-bubble interaction of wall-sliding bubbles is captured by using Particle Tracking Velocimetry. In this report, trial studies to introduce the bubble-bubble interaction into Eulerian-Lagrangian model have been performed. Firstly, the bubble-bubble interaction force vector is estimated by inverse analysis from the experimental results. The translational motion equation of a bubble is used to derive the interaction force vector. Secondly, two kinds of interaction models are proposed and applied to the numerical simulation. Comparison between the simulated and experimental results revealed that the introduction of the bubble-bubble interaction model enabled to predict much more accurate structure of sliding bubble motion than previous Eulerian-Lagrangian model.

抄録全体を表示

Shock wave propagation and collapse of gas bubbles caused by a shock in a water containing gas bubbles are of large interest in relation to characteristics of water hammer, blowdown of high pressure liquids in pipes, damping of dynamic influence of shock waves on obstacles, etc. The large-amplitude shock wave propagations in the bubbly liquid are investigated using a two-phase shock tube. Effects of incident shock wave strength and initial void fraction on shock wave pressure characteristics are clarified for incident shock wave strength ranging from 0.06 MPa to 0.25 MPa. Behavior of the bubbles collapsing behind large-amplitude shock waves is also clarified by pressure history measurements and high speed photography.

抄録全体を表示

A DSMC (direct simulation Monte Carlo) simulation using Dynamic Molecular Collision (DMC) model based on Molecular Dynamics (MD) calculation is applied for solving the two-dimensional nonequilibrium hypersonic rarefied flow over a flat plate with angle of leading edge. Numerical results show that nonequilibrium between translational and rotational temperature are obtained behind the leading edge over the plate. Translational and rotational energy distributions in the nonequilibrium flow show a typical distribution which is significantly different from the Maxwell-Boltzmann distribution and reveals a transition process from the nonequilibrium state to the equilibrium state.

抄録全体を表示

A DSMC (direct simulation Monte Carlo) simulation using Dynamic Molecular Collision (DMC) model and Multi-Stage (MS) model based on Molecular Dynamics (MD) calculation is applied for solving the two-dimensional nonequilibrium hypersonic rarefied flow over a flat plate with angle of leading edge. Numerical results show that nonequilibrium between translational and rotational temperature are obtained behind the leading edge over the plate. Pressure, heat flux ratio on the plate in the DSMC results agree well with those in the experimental results. The results also indicate that the effects of gas-surface interaction are revealed not only in the macroscopic states but also in the microscopic states.

抄録全体を表示

In the present paper, we propose a new method which is called HGA (Hybrid Grid Adaptation) method. It is very important to generate grids easily around complex objects in recent numerical simulations, in order to use limited computer resources efficiently. In the present paper, the kinds of grid, which have great effect on analysis accuracy and efficiency, are paid attention to. In twodimensional problems, we usually use only triangular and quadrilateral grid. The former, which has an advantage of freedom, can apply to complex abjects better than the latter. The latter can analyze flow problems more efficiently than the former. So we used hybrid grids to take the advantage of both freedom and efficiency. And we tried to analyze them more efficiently, using HGA method included AGR (Adaptive Grid Refinement) method, which obtains higher resolutions by collecting fine grids only in the location where the change of the physical quantity is intense, based on Hybrid Grid.

抄録全体を表示

A new approach to represent the non-slip boundary condition in the lattice-Boltzmann method (LBM) was proposed to improve the resolution for flow in the vicinity of solid wall. Considering the velocity distribution function as a volumetric average in each lattice, the solid boundary was taken into account in the convection process, not in the collision process as in the previous method. Then we applied it to the driven flow in a square cavity of Reynolds number up to 7 500. The wall-limiting profile of velocity and the numerical stability were significantly improved by our method in comparison with previous ones. We believe our proposal could contribute to develop an applicability of LBM for higher Reynolds number flows.

抄録全体を表示

Lift force on a spherical gas bubble rising near a wall in water was measured experimentally below Reynolds number (Re) of 20. We developed an experimental apparatus in which a CCD camera with a microscope follows the rising bubble. We used it to precisely measure the bubble radius, the rising speed and the distance between the bubble and the wall (L) and then estimate the lift force for Re<20. We also proposed the equation for estimating the lift force by modifying the analytical solution obtained under the Oseen approximation and compared with the experimental results. The results reveal that the lift force can be expressed by the analytical solution for Re<1. However, in the regime of Re>1, the experimental values are larger than the analytical values and the lift force should be expressed by a function of Re and L^*. We also proposed a predicting equation for the lift force as a function of Re and L/R by fitting the experimental data.

抄録全体を表示

When the Reynolds number is larger than the critical value or laminar separation is eliminated by a trip installed on a hemispherical cylindrical body, attached spot cavitation is observed occasionally. It occurs at fixed place in the vicinity of the minimum mean pressure and grows into a triangular wedge. In the present investigation, isolated and parallel spot cavitation is artificially generated on a hemispherical body and the behavior of the cavitation is observed by using high-speed movies and instantaneous photographs. The frequency of damaging blows in the range of the flow speed of 25 to 50 m/s is obtained by counting the number of pits on an aluminum specimen. When spot cavitation occurs adjacently, the cavities become rather stable. The pitting rate at the maximum damage zone by the parallel spot cavitation is much smaller than that by the isolated spot cavitation. The total pitting rats at the maximum damage zone by the isolated and parallel spot cavitation vary roughly 5th power of the flow speed for the both cases.

抄録全体を表示

In control valve technology, life time due to cavitation erosion is one of the most important items to be predicted. As the erosion develops, the throttle characteristics deviates from the original, and the seat leakage at the shut-down position increases. Even a rough estimation of the life-time in relation with on-cite history is necessary for a reliable operation of respective fluid system. In the present study, a contoured plug of stainless steel inserted in a globe valve has been subjected to erosion tests using a 20 MPa water test bench. Numerical analysis of cavitation-free flow and cavitation observation in a low differential pressure range were also made to obtain a fundamental insight of the valve flow structure. The erosion initiates from a stage of pit formation on the seat-joint-taper of the plug surface. Then, as the second stage, the weight loss increases rapidy with time as a process in which the erosion extends inward from the surface. The duration of the initial stage becomes short as the plug lift increases. The weight loss rate, which takes the highest at the full lift, is correlated by the kinetic energy of the valve throat flow. The weight loss in the second stage of erosion is also correlated by a power law to the working duration. The valve life time may be predicted by the correlation relationships based on the seat-leakage flow rate which is related to the erosion weight loss.

抄録全体を表示

When the propagating compression wave reaches at the open-end of a tube, the impulsive wave discharges to the surrounding area. The objective of this work is to investigate the usefulness of passive control using the modifications of tube exit for augmentation of the impulsive wave. The numerical simulations were carried out in order to clarify the effect of flare located at tube exit on the magnitude of the impulsive wave. The results were compared with those of the straight tube based on the aero-acoustic theory with the open-end correction and showed that the present technique of passive control enable to augment the magnitude of the impulsive wave.

抄録全体を表示

This paper reports about experimental investigations of the rotordynamic fluid force on a centrifugal impeller with three types of wear-ring seals; i.e., a face seal and two types of toothed seals. The impeller is equipped with a vaneless diffuser. Rotordynamic fluid forces on the impeller in whirling motion were measured directly by using a 4-axis force sensor. Unsteady pressures were measured at several locations in the diffuser. It was found that, (1) At low flow rate, the fluid force and fluid force moment become maximum at a certain whirling speed caused by a coupling between the whirl and vaneless diffuser rotating stall. (2) The seal geometry affects the direction of the coupled fluid force relative to the direction of eccentricity through the change of unsteady leakage flow due to the whirl.

抄録全体を表示

In this paper we describe the flow around two-dimensional circular cylinder bundles supported by wires in the staggered arrangement. The experiment was carried out in an N. P. L. blow-down-type wind-tunnel with a working section of 400 mm×400 mm×2 000 mm, and with the Reynolds number of 1.2×10⁴. The displacement, time-mean pressure distributions, and fluctuating pressure distributions of the oscillating circular cylinder were measured. The results were compared with those for the cases of the square arrangement. Consequently it was found that (i) while the displacement amplitude in X direction is nearly equal to that in Y direction, the occurrence of the fluid-elastic vibration is due to the hybrid phenomenon of jet switch and wake switch (ii) the variation of fluctuating drag coefficient and fluctuating lift coefficient corresponds to that of displacement of an oscillating cylinder, and increase rapidly with the occurrence of the vibration.

抄録全体を表示

Large eddy simulation of the complete stage of a mixed-flow pump is presented. The computation takes full account of the interaction between the rotating impeller and the stationary casing by using a multi-frame-of-reference dynamic overset grid approach. A streamline-upwind finite element formulation with second-order accuracy both in time and space is used for discretizing the governing equations. It is implemented in parallel by a domain decomposition programming model, and therefore, it can be applied to a large scale computation on a distributed-memory parallel computer. Internal flow of a high-specific-speed mixed-flow pump stage (ω_s=2.1) is simulated to test the validity of the proposed method under four operating conditions: the design point (Q/Q_d=100%) and three off-design points (Q/Q_d=60%, 80% and 120%). For all the cases computed herein, the predicted total pump head and the unsteady fluid forces on the impeller agreed well with the measured values.

抄録全体を表示

The vorticity-stream function formulation and moving overlapping meshes are applied to simulate the flow in the lobe pump. The stream function on rigid boundaries are determined by using the pressure single-valuedness and pressure gradient condition. As results, the pressure on the fluid machinery is given by using pressure gradient condition in the vorticity-stream function formulation. The present results show that these conditions are accurately satisfied, the feature of the discharge on the lobe pump is caught, and the flow is adequately simulated by these condition.

抄録全体を表示

The purpose of the present study is to develop micromachines imitating flagellated micro-organisms. The speed of advance, the radius of trajectory and the power consumption of a micromachine are computed by using the boundary element method to optimize seven parameters that express the shape of an actual microorganism and six parameters that stand for the flagellum shape of a micromachine. The optimum cross-sectional shape of the flagellum is an ellipse which is somewhat inclined with respect to the principal normal line of the flagellum spiral. The optimum thickness distribution of the flagellum is such that the position of maximum thickness is very close to the head of the machine. The optimization leads to the reduction of power consumption by 22 percent compared with the power consumed by the actual microorganism.

抄録全体を表示

The single-phase forced-oscillating-flow heat-pipes utilize the extraordinarily enhanced diffusion in oscillating flows inside channels. Dream pipes and COSMOS heat pipes are typical examples of such heat pipes and they have great potential for micro heat pipes and flexible heat pipes. On the other hand, they need external supply of power and so the optimization of system parameters is of great importance to reduce the power consumption. In the present paper, the optimum conditions achieving the highest effective thermal conductivity and/or the highest operating coefficient are analyzed for a given amplitude of the oscillating flow. The parameters in the optimization are the thermophysical properties of the operating liquid, the channel size and the frequency of oscillating flow. It is found that, under the optimized condition, COSMOS heat pipes excel dream pipes in thermal performance and ρc_p together with Pr is important to determine the optimum operating liquid.

抄録全体を表示

In order to investigate effects of adhered molecules and internal motion of oxygen molecules on reaction probability and energy transfer to metallic surface, we carried out a classical molecular dynamics simulation by using LEPS potential energy surface between an oxygen molecule and Ag surface. Reaction probability of oxygen molecules much depends on internal molecular motion, such as vibrational temperature and rotational temperature of oxygen molecules as well as coverage ratio of adhered molecules on the surface. Reaction probability averaged over incident angle of oxygen molecules has linear relation to the coverage ratio of adhered molecules on the surface. As a result, energy transfer to Ag surface averaged over incident angle of oxygen molecules has linear relation to the coverage ratio only when the coverage ratio is small.

抄録全体を表示

An experimental study was performed to investigate the flow and heat transfer around a square block, simulating a package of electronic equipment. The height and the side length of the block were H=2 mm and L=45 mm, respectively. The mean velocity of air in the rectangular duct, U_m, ranged from 0.24 to 4 m/s, and the duct height, H_D, ranged from 3.8 to 10mm. The thickness of laminar boundary layer was greater than the block height. It was found that the heat transfer distribution on the top face of the block was approximately correlated to the distribution based on the theory of laminar flat boundary layer. The average Nusselt number on the top face was correlated by Nu_m=0.61Pr¹/3(Re^*_L)⁰.5, where Re^* is the modified Reynolds number considering the effect of the opening ratio. This correlation was confirmed by many previous researches, performed under the similar condition of flow field.

抄録全体を表示

The effects of cross sectional shape of non-circular fibers on the radiative properties, such as extinction and scattering efficiencies, are numerically investigated. A finite element method is developed to solve Maxwell's equations that govern a 2-dimensional electric field induced by a single fiber that is normally irradiated by a plane wave. The angular profile of the scattered radiation as well as the scattering and extinction efficiencies are derived from the electric fields. The radiative properties of fibers with equilateral triangular cross sectional areas are compared with those of circular fibers. The results show that the intensity distribution of the scattered radiation are scarcely affected by the cross sectional shape when the size parameter is less than 2, and that the efficiencies of the triangular shaped fibers are equivalent to those of the circular fibers having the same cross sectional area. This study also shows the scattering efficiencies of fibers having rectangular cross sectional shapes become much different from those of the circular fibers as the aspect ratio of the cross section increases more than three.

抄録全体を表示

The stability of a circular liquid jet with near-critical mixing surface is analyzed in the frame-work of inviscid linear stability theory. The phase equilibrium condition results in vanishing small surface tension and surface density difference at high pressures. Because of very low mass diffusivity, the density changes only in the vicinity of the near-critical mixing surface. The analysis shows that the small surface density difference enhances the instability to a great degree, coupling with the effects of gas-phase velocity boundary layer flow and baroclinic vorticity production.

抄録全体を表示

Flow reactor experiment has been performed to study the promotion effect of fuel species on the NO-NO₂ conversion at reaction temperatures from 600 K to 1100 K. As the fuel species, n-butane (n-C₄H₁0) and n-pentane (n-C₅H₁2) were selected. The concentrations of NO, NO₂, and fuel species in the flow reactor were measured using a reacting flow of air/NO (20 ppm) /fuel species (10 ppm) mixture at residence times up to 1.5 s. The dependence of the NO₂/NO_X ratios on the type of fuel species, the reaction temperature and residence time are discussed. The experimental results show that both butane and pentane strongly promote the NO-NO₂ conversion even at the low reaction temperature (650 K). The most interesting results is the finding of the two peaks of the NO₂/NO_X ratios at lower and higher tempeartures. The appearance of the two peaks is thought to be due to the negative temperature coefficient of low temperature oxidation reaction for higher hydrocarbons.

抄録全体を表示

Basic considerations to identify dioxins and related compounds formation through Municipal Refuse Incineration process are classified into two different fields. One is concerned with combustion process another with after-combustion process. For the after-combustion field, dioxins formation theory is well established with apparent hack data. This paper describes the first attempt to verify the dioxins formation pathways, which are predicted as the combination of HCl, Cl₂, Aromatic Chloride Compounds, and Dioxins (by-products) formation in combustion zone, through basic chemical reactions and small scale combustion experiments.

抄録全体を表示

An experimental study was made on the structure of the turbulent counter flow non-premixed flame established in the forward stagnation region of a porous cylinder. Stable-species-concentration profiles were measured in detail for propane flames at atmospheric pressure. These distributions were analyzed to make clear the effects of flame stretch and turbulence intensity on the flame structure. The non-premixed flames are extinguished by flame stretch caused by the flow field. For the turbulent flow, effects of the flow field can be separated into those of the bulk flow and fluctuating flow. Therefore, the flame stretch consists of the bulk flow stretch and turbulent stretch. In the present study, the turbulent stretch was estimated by the inverse of the Kolmogorov time scale. With increasing the flame stretch, thickness of the turbulent flame zone decreased and spatial gradient of the species become steeper. The main reaction generating the intermediate species occurred on the fuel side of the turbulent flame zone. With the strong stretch, leakage of the intermediate species is large compared to the weak stretch. With the strong turbulence, the profiles of the intermediate species are rather flat, whereas the Gaussian distributions are observed for the case of weak turbulence. Based on the mixture fraction, the effects of flame stretch and turbulence intensity cannot be discernible.

抄録全体を表示

A method is developed to estimate asymmetrical flame structure (profiles of temperature, soot density and CO₂ partial pressure) in flame containing soot from transmission and emission data. This method considers self-absorption of radiation by CO₂ and soot. It needs radiation measurements at two wavelengths. One is the wavelength at which monochromatic absorptivity of CO₂ is almost 0 and only soot absorbs radiation. The other is the wavelength in which both CO₂ and soot absorb radiation. Using these radiation data, flame structure can be reconstructed. In this paper, using the radiation data simulated from assumed flame structure, flame structure was reconstructed. When the distribution of temperature was assumed to be uniform and random error within ±0.5% is added to the simulated radiation data, standard deviation of the estimated flame temperature is 6.3 K.

抄録全体を表示

Premixed lean diesel combustion (PREDIC) achieves very low NO_x emissions as a result of early fuel injection. Another consequence of such early injection is an increase in the amount of fuel spray reaching the cylinder wall. This results in higher fuel consumption and THC emissions. Another shortfall is that compression ignition timing control is difficult because of its dependence upon the incylinder temperature. These problems can be addressed by utilizing an impingement spray, with two nozzles. In this study, this type of spray was analyzed on spray position control, larger spray volume, and less wall wetting, among others. Also, the impingement spray structure was analyzed by cross-sectional observation using laser induced fluorescence. Spray penetration, high-density area, and plume area were measured. The impinging sprays enhance each other's spreading and mixture formation. Higher injection pressure has little effect on spray penetration and high-density area.

抄録全体を表示

Recently, pre-mixed lean combustion has gained the attention of many researchers, who are investigating compression ignition characteristics of this type of mixture. Previous research in our laboratory has shown that low NO_x combustion can be obtained by PREDIC (PREmixed lean DIesel Combustion). While NO_x emissions were very good, the fuel consumption was greater than that of a conventional engine, due to a lack of ignition timing control. So in this study, ignition timing control was attempted using two fuels. Results showed that ignition timing control was possible using a low cetane fuel (such as methanol or α-methyl naphthalene) after the diesel fuel main injection. The injection timing and quantity of methanol affected the ignition timing of PREDIC.

抄録全体を表示