This paper describes a multi-step strong coupling method to analyze an equation system of incompressible viscous fluid and an elastic body. The procedure is derived based on the following techniques; (a) sub-structuring, (b) velocity-pressure splitting, and (c) time integration of Newmark's family. Only simultaneous equations with well-defined symmetric matrices are solved in the method. To demonstrate fundamental performance of the method, it is applied to solve an impacted beam in fluid. The result shows good agreement with a previous work.

抄録全体を表示

Three-dimensional vortical flow structures and velocity fluctuation near the rotor tip in an axial flow fan have been investigated by experimental analysis using a rotating hot wire probe and a numerical simulation. Both the experimental and the numerical analysis have been performed in the relative frame of reference rotating with the rotor. It is found that a tip leakage vortex is observed in the blade passage, which has a major role near the rotor tip. The tip leakage vortex formed close to the leading edge of the blade tip on suction side grows in the streamwise direction, and forms a local recirculation region resulting from a vortex breakdown inside the blade passage. It causes significant changes in the nature of the tip leakage vortex: large expansion of the vortex and large total pressure loss. High velocity fluctuation is observed in the interference region between the tip leakage vortex and the main flow. Near the casing wall, a discrete frequency is formed between tip leakage vortex center and rotor trailing edge.

抄録全体を表示

Measurements of the mean and the turbulent velocities have been performed in the vicinity of a roughness element for the turbulent boundary layer developing over a k-type rough wall, which consists of two-dimensional square ribs arrayed with pitch ratio of 4. The Reynolds-numbers based on momentum thickness θ and roughness height k are about 500 and 50 respectively, and relative roughness height k/δ6=0.156. LDV (Laser Doppler Velocitimetry) that it is capable to measure reverse flow was used in order to clarify the momentum-exchange process in the open-face (y_T=0 mm) between roughness elements and the flow-pattern in a groove. The variation of the mean and the turbulent quantities profiles at Y_T=0 mm depends on the development of a free-shear flow formed with disappearance of the roughness element and the formation of eddies in the groove. Pressure drag coefficient acting on the roughness element in the local skin friction coefficient is 92% in the present k-type rough wall. The momentum exchange for the mean flow and turbulence contribute 6% and 59% respectively to the local skin friction coefficient. The existence of a large-scale eddy and two small corner-eddies is confirmed from the flow pattern in the groove.

抄録全体を表示

A standing wave with a peak line toward downstream appears on the surface of liquid film flowing around a horizontal circular cylinder. The amplitude of the wave grows in the downward direction since the centrifugal force grows because of the acceleration of the film flow by gravity. An equation of motion governing the wave amplification was theoretically derived. The amplitudes calculated by the equation agree with those measured experimentally for three kinds of test liquids. The approximate consideration of the above equation shows that the amplification ratio A is mainly dependent on a parameter Fr₀(R/δ₀) concerning gravitational and centrifugal forces, and slightly on the Reynolds number. Since the surface tension acting on the wave is linear to the film thickness, its effect on the amplification can be neglected.

抄録全体を表示

The behavior of the boundary layer flows in rotating channels such as the flow in impeller passages of turbomachinery is dominantly influenced by the centrifugal and Coriolis forces due to the wall curvature and system rotation, respectively, which predominate the boundary layer development by suppressing or promoting the turbulent motion. Though many studies have been perfomed about the turbulent boundary layer over a concave or convex surface in the stationary state, the resultant effects due to the Coriolis force in the rotational frame of reference have been scarcely researched. In the present study the velocity and turbulent intensity are measured near the concave surface in two different channels whose radii of center line curvature are 1000 mm and 2000 mm, respectively, on the rotating system. When the Coriolis force acts toward the concave wall due to the channel rotation, the transition of the boundary layer to the turbulent state is promoted because the Gortler instability is enhanced. For the prediction of the boundary layer transition, Gortler number can be an appropriate parameter because the transition occurs at the point where Gortler number exceeds 7, irrespective of the axial Reynolds number, rotation rate, or curvature radius of the wall.

抄録全体を表示

Analysis of compressible gas-liquid two-phase flow is important in various industrial fields. A new numerical method of the compressible two-fluid model is developed. An approximate treatment of the Riemann invariance, which is necessary to treat the relationship between discontinuities of pressure and velocity in highly compressible flow, is introduced to the two-fluid model calculation. Two-phase volumetric velocity and mixture sound speed are adopted in the Godunov scheme for solving the Riemann invariance in two-phase flow. This technique is employed in the finite volume method for two-fluid model calculations and the calculated speed of sound is in good agreement with the theoretical one. The present method is applied to analysis of experiments of pressure wave propagation in a pipe filled with water and two-phase mixture. Predicted pressure responses from axial one dimentional analysis in the pipe are in good agreement with the experimental data.

抄録全体を表示

A numerical analysis was carried out to clarify the unsteady behavior and the large-scale structure of high-speed submerged water jet with cavitation. A finite volume formulation with an upwind scheme was used to discretize the compressible Navier-Stokes equations with an assumption of locally homogeneous model of gas-liquid two-phase media. First, the validation of the present numerical method was discussed by comparing numerical results with experiments of jet flow. Further, the difference between the jet structures with and without cavitation was investigated. As a result, it is indicated that the cavitating jet flow corresponds basically to an unsteady transonic flow condition. Also, it is found that the present numerical method is available to predict the complex structure and the unsteady behavior of the cavitating jet flow.

抄録全体を表示

The passive flow control was numerically studied for a two-dimensional rectangular cylinder. Several control means were proposed so far like a small obstacle placed in front of a rectangular cylinder. One of the present authors proposed to connect a front obstacle and a rectangular cylinder with a splitter plate. The controlled flow field was investigated by the Navier-Stokes equations with the modified LES turbulent model. The results were compared to a previous method and it was found that the improvements were achieved in the drag reduction.

抄録全体を表示

Recently, numerical simulations of two-component fluid fow is performed with highperformance computer. In these method, numerical solution is given by solving a set of partialdiffererencial equation modeled with experiments. However numerical divagency is sometimes occurred by complexity of two-component interface. Moreover, it is very difficult to evaluate interfacilal area cocentration with experiments. The 2-dimiensional immiscible lattice gas (ILG) model of lattice gas automata method is expected to describe complex interface behavior. In the present study, numerical simulation of two-component fluid fow is peformed with 3-dimensional ILG model. And the applicability of this model to two-component fluid flow is validted.

抄録全体を表示

The purpose of the present study is to develop a new type snow fence with induced snow plate in the form of circular arc airfoil for the prevention of snowdrifts and improving visibility on roadways. The design is in accordance with aerodynamic principles, and its performance is evaluated in a wind tunnel in which natural snow is used to simulate blowing-snow. In general, the snow fences are classified into two types, i.e., snowbreak type and snow blowing-off type. The developed snow fence in present study has two performances of which blowing-snow is broken at the bottom part and is blowed from the top part of snow fence. An optimum configuration of the induced snow plate which forms the top part of the snow fence is designed for the prevention of blowing snow at the leeward side on roadway. Consequently, the developed new type snow fence with the induced plate in the form of circular arc airfoil has high performance in preventing snowdrifts and improving visibility in comparison with the conventional snowbreak fence. In especial, the performance of the newly developed snow fence is in most cases preserved without burying itself in snowdrifts which often occurs in the conventional snow fence when a large-scale blowing-snow occurs.

抄録全体を表示

Flow characteristics around the D-shape cylinders which cut out the front face of a circular cylinder and the I-shape one which cut off the front face and rear faces were investigated in the range Re=(1.38∼8.0) × 10⁴. The following conclusions were obtained. For the D-shape cylinder with the cutting angle of 53 deg at Re ≥ 2.3 × 10⁴, the shear layer separated from the front edges reattaches on the circular arc of the cylinder, and the transition in the boundary layer and the turbulent separation occur. As the results, the wake width decreases and the vortex formation region goes downstream. The Strouhal number increases beyond 0.28 and the base pressure coefficient has a maximum of -0.70. The drag coefficient of the D-shape cylinder decreases to 0.7. The same phenomenon occurs for the I-shape cylinder with the front cutting angle of 53 deg and the rear cutting angle of 115 to 127 deg at Re ≥ 2.3 × 10⁴.

抄録全体を表示

The deformation of the interface between a water drop and silicone-oil surface impacted by it was examined in detail. Special attention was directed to the influences of the water-drop diameter d_L on the deformation of oil surface and on that of the water drop itself after the impact with the oil surface. The drop diameter was varied from 1.9 mm to 4.1 mm. The influences of d_L on the shape of oil cavity at its maximum depth and on the drop shape at its maximum spread were observed conspicuously in case of minimum oil viscosity of v_T=5 mm²/s. It was also found that both the maximum cavity depth ratio D_M/d_L and the maximum diametral deformation ratio d_M/d_L became smaller as d_L was increased. These deformation parameters D_M/d_L and d_M/d_L could be well correlated with the impact velocity of drop u_L for each oil viscosity. Moreover, in Case B, C and D, D_M/d_L could be correlated by a dimensionless group Re_TLWe_TL, and d_M/d_L was correlated by Re²_LOh for highly viscous silicone oil of v_T ≥ 5 × 10³ mm²/s.

抄録全体を表示

In order to develop a phase separation technique under microgravity condition, the method of the impacting Y-junction by utilizing the difference of inertia force between gas and liquid phases is proposed. Experiments of phase separation in an impacting Y-junction of 10 mm ID placed in a horizontal plane were carried out under normal and microgravity conditions using air or gaseous nitrogen and water as working fluids. The phase separation performance under microgravity was measured by controlling the extraction rate during the microgravity period of about 20 seconds by parabolic trajectory flight of the airplane (G II). As a result, it was shown that liquid extraction from two-phase flow was possible by using a Y-junction for every condition in the microgravity experiments. About 60% maximum of all the liquid phase could be extracted from incoming gas-liquid two-phase flow by Y-junction under microgravity in our experimental range.

抄録全体を表示

An initial lift force exerted by a quiescent fluid on a circular cylinder suddenly starting and rotating from rest is much complicated phenomena, because it results from the nonlinearity between the rotating and the translating motions. In the present paper, the lift coefficient, C_L, at the extremely early stage of motion is analytically obtained by using the matched asymptotic expansion method. The previous paper⁽1) showed that C_L for time T→0 exhibited the T^-1/2 singularity at the low Reynolds number (Re<< 1), however, the present paper shows that C_L becomes asymptotically constant at Re=О (1), which is different from the theoretical prediction (C_L∼T¹/2) given by Badr & Dennis⁽2) at Re >>1. Furthermore, the present paper shows that the dominant term of the initial drag coefficient, C_D, is the T^-1/2 singularity at three cases; Re<< 1, Re=О(1), and Re >>1.

抄録全体を表示

It is known that the long time scale fluctuation about 20 times the vortex shedding periodicity exists in the irregularity of the vortex shedding from a two-dimensional circular cylinder. However it's origination is not known yet. Local vortex shedding frequency in spanwise direction is considered to differ from one point to another due to different local Strouhal number caused by flow three-dimensionality. Therefore, interaction between the different vortex shedding frequencies is considered to be the cause of the long time scale fluctuation. In this study, by means of short time Fourier transformation which is modified to have uniform local time response characteristics throughout entire measuring frequency range, modulation characteristics of quasi-instantaneous vortex shedding frequency observed by one point velocity measurement are investigated. Moreover, by means of the modified short time Fourier transformation, two point simultaneous velocity measurements were made to show the existence of the different vortex shedding frequencies at the same time in the spanwise direction.

抄録全体を表示

We have in vestigated the rheological properties and the orientational distributions of particles of a non-dilute colloidal dispersion, which is composed of ferromagnetic spherocylinder particles, subject to a simple shear flow. The mean field approximation is applied to take into account the interactions between spherocylinder particles. The basic equation of the orientational distribution function has been derived from the balance of the torques (including the term due to the mean field approximation) acting on the particle in an applied magnetic field. It is an integrao-differential equation. Then, the governing equation has been solved by means of the method of successive approximation and Galerkin's method. The results obtained here are summarized as follows. For the case of strong magnetic interactions between particles, the particle exhibits a sharp peak of the orientational distribution even for a weak applied magnetic field. In this case, the average magnetic moment of the particle becomes large, which leads to strong interactions between the applied magnetic field and the particle. Thus, the particle tends to point to the magnetic field direction under these situations. Also, in this case, the large increase in the viscosity is obtained due to such a restriction concerning the particle orientation.

抄録全体を表示

A new subgrid-scale (SGS) model for practical large eddy simulation (LES) is proposed. The model is constructed with the concept of mixed (or hybrid) time scale, which enables us to use the consistent model parameters and to dispense with the distance from the wall. The model performance is tested in plane chanel flows, and the results show that this model is able to account for nearwall turbulence without an explicit damping function as in the dynamic Smagorinsky model. The proposed model is also applied to the backward-facing step flow examined by Kasagi and Matsunaga(1995) experimentally. The calculated results show good agreement with experimental data, while the results obtained by using the dynamic Smagorinsky model show less accuracy and less computational stability. These results suggest that the present model is a refined SGS model suited for practical LES to compute flows in a complex geometry.

抄録全体を表示

A novel laser Doppler velocimeter (LDV) technique which uses two single-frequency diode lasers, one for each LDV beam, is presented. The existing techniques using the optical frequency difference between two lasers for directional discrimination was enhanced by the technique discribed here. The Doppler frequency extracted by an optical superposition technique and a pair of quadrature signals in the base band was generated without any broadband electronic heterodyning. The electronic heterodyning on the conventional systems was implemented by an electronic mixing technique of the beat signal of the two lasers and the measuring signal generated by the scattered light of the LDV beams. This paper describes a new optical technique in contrast to these electronic heterodyning systems. In order to extract the Doppler frequency optically, undesired mixing products must be avoided. The realization of this concept is based on the superposition of four light waves on one photodetector, where only two pairs of light waves are made to forcedly interfere by polarization or by optical adjustment. Thus, the photodetector acts as an optoelectronic element that superimposes two optical beat signals, so that the Doppler frequency is obtained from the envelope of the resulting output signal directly. In order to obtain the desired directional information, a second photodetector for quadrature signal generation is required. The phase delay between the detected signals was fixed by positioning in the expanded wavefront of the reference beams. A prototype of the directional LDV, which uses two lasers and the described optical mixing, has been realized.

抄録全体を表示

A purpose of this study is that unsteady exaluation of thermal performance of heat exchanger with frosting and decision of optimal defrosting cycle. Because increase of flow resistance is the principal factor of a drop of heat transfer performances with frosting, thermal performance characteristics were examined. Based on that experimental result, we proposed one-dimensional non-homogeneous frost growth model, and compared with the experimental data in time and space. In a non-homogeneous model with the frost property distribution taken into account, it shows that density similar to the experimental result can be predicted. It is possible to make prediction closer to the experimental result as compared with the conventional homogeneous model with respect to the time variation of frost height.

抄録全体を表示

Fundamental properties of the process of heat exchange between atmospheric pressure non-equilibrium plasma and solid wall have been inventigated by focusing on a surface discharge. The surface discharge was exposed to air and nitrogen jet impingement to secure the flow field as well as feed gas supply, and in situ measurement of the wall temperature was employed by thermography. About 80% of the discharge power was measure as heat flow to the wall and heat transfer rate was dominated by the generation of Joule heat. On the other hand reaction enthalpy or recombination of electric charges showed minor effect on heat transfer rate due to the low degree of ionization. Such Joule loss was successfully suppressed by 50% with the fast rising pulsed voltage, which is well-known technique to minimize ion current, thus heat flow rate to the insulating wall could be also halved compared to the results obtained with the high frequency sinusoidal voltage (10 kHz-80 kHz).

抄録全体を表示

This study proposes the turbulent kinetic-energy equation to describe the behaviors of the production and the dissipation in porous media. We clarify the mechanism of the production and the dissipation intrinsic to the turbulence in porous media. The production characterizes that the turbulence is produced by the vorticity which is induced by the solid obstruction. On the other hand, the dissipation characterizes that the obstruction of a solid matrix dissipates the eddies, which tend to become larger than its representative length, to the interstitial eddies. Furthermore, the Kolmogorov's length scale for the turbulent flow through porous media is obtained from the order estimate for the proposed turbulent kinetic-energy equation.

抄録全体を表示

The present experiment and numerical analyses have been performed for obtaining the cold heat storage characteristics of a water mixture of microcapsules packed with liquid-solid phase change material (PCM) flowing in a pipe cooled under the constant temperature wall condition. The energy equation was formulated by taking into consideration of the heat release due to the PCM solidification process and the heat transfer enhancement induced by the microconvection of microcapsules. The heat source function in the energy equation was derived from solutions for the solidification in a spherical latent heat material. The governing parameters were found to be latent heat material concentration, cooling wall temperature, microcapsule diameter and velocity of water mixture of the microcapsules. The experimental and numerical results revealed that mean heat transfer coefficient for the water mixture of the latent microcapsules was about 1.6∼2.5 times greater than that for the single phase of water.

抄録全体を表示

Characteristics of the multistage high pressure reciprocating compressor, which expects to obtain outlet pressure over 30 MPa, have been studied. For improvement of volumetric efficiency, the performance characteristics are comprehensively studied by comparing the numerical results with those of experiments. These measures of pressure pulsation, valve dynamic behaviors and flow rate have been executed by experiment. Using the numerical simulation code of STAR-CD, which assumed three dimensional, compressible and unsteady flow, the dynamic phenomena of compressor parts has been demonstrated. Consequently, parameters that need for improvement of flows in the compressor have been discussed.

抄録全体を表示

The effects of oxygen on coal gasification characteristics have been investigated by using Drop Tube Furnace. In the pyrolysis tests the influences of oxygen in the oxidizer have been examined, and results show that the reactions in which oxygen don't take part isn't influenced by pressure. The gasification tests with changing oxygen fraction in oxidizer have been done, and results show that, for the high volatile matter coal gasification efficiency isn't influenced by oxygen fraction, but for the low volatile matter coal, it is improved by increasing the oxygen fraction. Moreover, the reactivity of the low volatile matter coal has been investigated, and the effects of oxygen fraction are clarified by comparing the effects of the temperature and the residence time of coal particle.

抄録全体を表示

Devolatilization behavior of pulverized coal has been observed with flash pyrolizer. Produced gases emission behavior has been investigated by high-speed continuous analysis with FTIR, and so on. The following results are acquired. (1) Each emission gas takes a characteristic emission behavior, and the hydrocarbon influences the emission behavior of CO and CO₂. (2) All yield of each emission gas increases by increasing in amount of volatile matter in coal for five kinds of gas analyzed with FTIR, but by decreasing in H₂. (3) The difference of the yield of volatile matter by the heating temperature is hardly seen. (4) The ratio for the tar in volatile matter decreases by decreasing in volatile matter. (5) The difference of the emission time of the produced gases by kind of coal is hardly seen.

抄録全体を表示

Numerical calculations are conducted to explore the way how a cold droplet is ignited by an approaching diffusion flame and to characterize the three flame propagation modes classified in the first report. In the calculations, the convective terms are neglected and an Arrhenius type of one-step overall reaction is considered. When the flammable gas mixture layer formed around the droplet is reached by the diffusion flame, a premixed flame is formed in front of the diffusion flame and propagates in the layer. Since the period in which the premixed flame burns the fuel vapor out is very short, the droplet temperature does not change significantly during the period. Thus, a spherical diffusion flame is newly established around the cold droplet. When there is no flammable gas mixture around the cold droplet, the diffusion flame passes the droplet and the droplet comes to a member which supplies fuel vapor to the diffusion flame.

抄録全体を表示

The largest exergy loss comes from the combustion processes in the combustion systems such as gas turbines. The flow with the chemical reactions involves four irreversible processes of the viscosity, heat conduction, mass diffusion and chemical reaction. The source of the exergy loss during the combustion can be evaluated by analyzing the entropy generation of each process. In this study, we analyzed the entropy generation of four processes and the exergy loss in the one-dimensional laminar premixed flames. It is noted that the chemical reactions are responsible for the major exergy loss. The contribution of the heat conduction is significant, but less than that of the chemical reaction. An increase in the inlet flow temperature reduces the exergy loss, while a decrease in the equivalence ratio increases the exergy loss.

抄録全体を表示

Previous research by the authors showed that hydrogen combustion exhibits a higher cooling loss to the combustion chamber wall of an internal combustion engine compared to hydrocarbon combustion because of its higher burning velocity and shorter quenching distance. The high cooling loss means that reduction of the cooling loss is essential to establish a high thermal efficiency in hydrogen combustion engines. This research analyzed the applicability of equations to describe the heat transfer from burning gases to hydrogen combustion. The result shows that equations calculate a lower cooling loss than experimental values, and the use of correction coefficients does not accurately define the actual cooling rate. It is therefore concluded that the derivation of a new heat transfer equation for hydrogen combustion is necessary to improve the thermal efficiency of hydrogen fuelled engines.

抄録全体を表示

Characteristics of the flame propagation into the lean region in the stratified mixture were investigated by the experiments and numerical simulations. In the experiments, gaseous propane was injected into the lean propane-air mixture charge, whose equivalence ratio was less than the lower flammability limit of the premixed mixture. The flame generated in the fuel jet propagated wide into the lean mixture charge region as the stoichiometry of the mixture charge was increased. Then the interactions between the rich flame and the lean mixture were investingated by the numerical simulations of counterflow premixed flame. Reaction rate of the fuel in the lean mixture depended on both opposing gas composition and temperature. These results show the flame in the rich region propagates into the lean region by the "flame inertia" of the rich flame until its relaxation.

抄録全体を表示

The planar solid oxide fuel cell (SOFC) with the Y₂O₃ stabilized ZrO₂ electrolyte is expected to be a candidate for the distributed power sources in the next generation due to its high efficiency of power generation. In this study, we have analyzed the system performance of the SOFC and gas turbine combined cycle of about 500 kW power output, using a two-dimensional simulation program for planar SOEC with the internal reforming. The effects of cell temperature, SOFC pressure, recirculation ratios of fuel and air, utilization ratios of fuel and air, and average current density at SOFC on the system efficiency were analyzed under typical operating condition taking account of realistic efficiencies and heat losses.

抄録全体を表示