Propagation of shock waves in dilute bubbly liquids is investigated numerically. Governing equations for the bubbly liquid are formulated with emphasis on the radial and transverse motions of the bubbles. The conservation equations for mass, momentum and energy inside the bubbles are solved directly in order to precisely estimate the effects of internal phenomena on bubble motion. A numerical method, where individual bubbles are traced to estimate the effects of volumetric changes and relative motion on wave phenomena, is developed. For a wave process at steady-state conditions, simple relations, such as the propagation velocity, C_i, are derived. Numerical results under several conditions reveal that the terminal wave propagation velocity coincides with the velocity in the equilibrium condition, C_i. The slippage between bubbles and liquid has some influence on the time evolution of propagation velocity of the shock wave and the relaxation structure behind the wave. However, the slippage plays a minor role in the wave propagation processes.

抄録全体を表示

Propagation of shock waves in dilute bubbly liquids is investigated numerically taking into account internal phenomena of the bubbles. In this numerical analysis, the conservation equations for mass, momentum and energy inside each bubbles in the bubbly liquids are solved directly, in order to precisely estimate the effects of internal processes on wave phenonema. Numerical results under several conditions reveal that the radial motion of bubbles, which is affected by internal phenomena, such as thermal conduction through the bubble wall, has a significant influence on the time evolution of the propagation velocity of the shock wave and the relaxation structure behind the wave. The nondimensional thermal diffusivity, D, defined in eq. (13), is one of the most effective parameters to be correlated with the wave propagation process. As the initial bubble radius becomes larger or the pressure ratio becomes smaller, the time required for developing shock waves to steady-state conditions increases and the first maximum of the pressure profile in steady-state conditions becomes lower.

抄録全体を表示

The diameter and the position of bubbles in the tube cross section of upward air-water bubbly flow were measured using the image-processing method on two back-lit images which were taken from directions at right angles to each other. The bubble configurations, such as void fraction, bubble elongation or bubble diameter aspect ratio profiles, were calculated from the data assuming the bubble shape to be an ellipsoid. Elongation of bubbles in the flow direction was observed near the wall in sliding bubbly flow. The measurement error of the image-processing method was examined. The image-processing method was confirmed to be useful under low void fraction conditions.

抄録全体を表示

Pseudoshocks in straight channel flows were numerically simulated by solving the two-dimensional Reynolds-averaged Navier-Stokes equations. The object of the present simulation was to investigate the influence of parameters on the structure of pseudoshocks and to validate the model of the pseudoshock mechanism proposed in the previous paper. In this simulation, Harten-Yee's second order-accuracy TVD scheme was used for sharp resolution of the shock wave and numerical stability, and Baldwin-Lomax's algebraic model was used for turbulent flows. The computational results agreed well with the experimental ones and with the proposed model of λ-and X-type pseudoshocks.

抄録全体を表示

In this paper the saltation velocity, that is, the fluid velocity at the point at which the total pressure drop per unit length of the solids-air two-phase flow becomes minimum, is examined from the viewpoint of low pressure drop and power consumption. An empirical equation to predict the saltation velocity is first derived on an established flow in a horizontal pipe. Second, the experiment on the saltation velocity of coarse particles for the flow was conducted. The foregoing equation is sufficiently accurate for the experimental data, and useful for the approximation as to any particles resembling those in the above experiment.

抄録全体を表示

In order to clarify the flow characteristics of two-phase flows in pipes under microgravity, general correlations on the volumetric fraction and the frictional pressure loss in the pipe were obtained from the theoretical analysis based on an annular flow model. They were compared with the experimental results which were obtained in an immiscible equal-density silicone oil-water two phase flow in a horizontal tube of 25 mm in I.D. and also in a saturated R114 two-phase flow under microgravity in an aircraft. As a result, the calculated results of the frictional pressure loss were found to agree well with the experimental results by taking the values of 10 to 100 as the constant B for interfacial friction in the silicone oil-water liquid-liquid two-phase flow and values B close to 10 in the saturated R114 two-phase flow.

抄録全体を表示

Organized structure in a periodically diverging-converging turbulent channel flow is investigated experimentally. Mean frequency and duration time of bursting are obtained using the VITA method. The contributions to uv in the four quadrants of the uv-plane are examined by means of the quadrant analysis technique. The bursting frequency and the contribution of the ejection to Reynolds shear stress decrease in the converging section and increase in the diverging section. These results are consistent with the finding that the Reynolds shear stress decreases due to acceleration and increases due to deceleration.

抄録全体を表示

The final objective of this study is to find a turbulence model suitable for high-Reynolds-number flow with large separation, which is based on the compressible Navier-Stoke's equation. As the first step in achieving this purpose, Jameson's artificial dissipation, Baldwin-Lomax's zero-equation turbulence model, and Myong-Kasagi's and Chien's K-ε turbulence models were investigated in this paper. The flows in a suddenly expanding pipe were chosen as the test problem. It is demonstrated that Chien's K-ε model is the best one in terms of accuracy and grid independence of the reattachment distance, and in terms of stability of computation.

抄録全体を表示

Effect of the Reynolds number on turbulent characteristics is studied, using the data reported so far for fully developed turbulent channel flows. The data show that the turbulent intensity and the Reynolds stress distributions normalized by the friction velocity depend on the Reynolds number Re* if Re* is not too high. If Re* exceeds a given value, the distributions approach asymtotic curves within the wall region. This means that there exists the law of the wall for turbulent quantities. The conditions under which the law holds differ for the mean velocity and the turbulent quantities. Regime diagrams of turbulence are presented, which clearly indicate the Reynolds number dependence of turbulent quantities. The Re* dependence is considered in relation to the turbulent velocity scale, inactive motion and the wall-blocking effect.

抄録全体を表示

A new variable-order finite difference method is proposed for highly accurate numerical simulation of turbulent flow. In this method, the spatial derivatives are discretized by the Taylor series expansion, and the expansion order is variable. In the present study, the central difference schemes are formed for an even expansion order, and the upwind difference schemes are formed for an odd one. The method is applied to the numerical simulation of two-dimensional isotropic turbulent flow. The numerical results are discussed on the basis of the energy spectra. They show that the present higher-and odd-order methods effectively simulate turbulent flow, and the -3rd power law of the inertial range energy spectra can be estimated.

抄録全体を表示

Using direct numerical simulation (DNS) database for wall turbulence, we have performed a critical assessment of existing dissipation-rate equations for second-order closures and proposed a new dissipation-rate equation which is an extension of our earlier k-ε modeling. The DNS data of mean velocity and turbulent kinetic energy are used as the exact known quantities for solving the ε-equation. Thus, the values of ε obtained are the pure solutions of each ε-equation. The assessment has revealed that all existing models provide inaccurate predictions of the exact ε profiles. The proposed model, on the other hand, shows good performance, in spite of simpler model formulation.

抄録全体を表示

In our previous work, a novel finite element scheme for solving the unsteady incompressible Navier-Stokes equations up to high Reynolds numbers was developed. This scheme is based on the Petrov-Galerkin formulation using exponential-type test functions. The numerical results obtained in the previous work demonstrate that the method is capable of solving the incompressible Navier-Stokes equations accurately and in a stable manner for a wide range of Reynolds numbers. In this paper, the Petrov-Galerkin finite element method using exponential functions is extended to natural convection in a two-dimensional enclosed square cavity. The unsteady incompressible Navier-Stokes equations and energy equation are discretized by a semi-explicit scheme with respect to the time variable. As the time-marching scheme, the fractional step method is used effectively in this study. The validity of the present method is shown for unsteady flows at high Rayleigh numbers through comparison with other existing numerical solutions.

抄録全体を表示

For the study of the combustion in a transient gas jet used in gas diesel engines, it is very important to investigate the entrainment process of the surroundings of the jet. In this study, a path line method utilizing a CCD camera with various shutter times was applied. Large (40μm), light weight micro-balloon particles are used for scattering the path lines of the surroundings and fine particles (0.75μm) are used for visualizing the approximate shape of the jet. The transient gas velocity was also measured with a two-component laser Doppler velocimeter. Consequently, (1) the jet shape and the entrainment process could be visualized simultaneously, and (2) this path line method was found to be very useful fot estimating the amount of entrainment air because of the good agreement with the value obtained using LDV data.

抄録全体を表示

Cellular automata have been used recently for simulations of fluid motion. However, applications of this method have been limited to flows with unchanged cinfigurations becase of difficulty with the boundary conditions. In this paper, we studied the flow around a flat plate moving in the normal direction in a rectangular domain by using the FHP-1 model, and presented a procedure for treating its boundary conditions. The results of the simulation agreed well with those of experiments of flow visualization.

抄録全体を表示

Normal forces generated in the flow between two-rolls, one rotating and the other fixed, are measured. Water and glycerin-water solutions are used in the experiments and the forces measured under the usual experimental condition are in good agreement with one-dimensional theoretical predictions obtained by Cameron using the Reynolds equation and the half Sommerfeld condition. A thin liquid film is produced on the surface of the rotating roll under the usual experimental conditions. But when the liquid film is stripped off with a spatula, the measured normal force is reduced. In the case of a small quantity of supplying fluids and the low speeds of rotating, the measured normal force is negative for water but positive for glycerin-water solutions.

抄録全体を表示

This paper presents numerical analyses for two-dimensional flow problems combining the finite element method, which is best fitted to flow analysis with complex geometry, and the finite difference method, which is economical in computing time and computer storage. Channel flows with a cavity are solved by means of the finite difference method, the FDM-FEM composite method, and the finite element method. Furthermore, the CPU time and memory size are compared. The results obtained show that the FDM-FEM composite method has both computational efficiency and geometrical flexibility. As an application to practical problems, flow in a combustion chamber is calculated.

抄録全体を表示

In order to gather information regarding the method of reducing aerodynamic noise of a high-speed elevator, the flow structure around an elevator car was studied by means of flow visualization experiments and numerical simulation using simplified models. From the study, it was made clear that flow behavior in front of the car door during downward motion was qualitatively different from that during upward motion, and that the difference was caused by the apron. During downward car motion, streamwise vortices are generated around the apron side edyes, following a mechanism similar to the generation of the tip vortex. As a result, the flow in front of the car door is converged and accelerated, and this mechanism causes the elevation of noise inside the car. On the other hand at the side walls and the back wall, large-scale separation occur and make the flow field very complex.

抄録全体を表示

It is very difficult to measure the flow patterns within rotating elements of a torque converter due to the complicated construction. In this study, three-dimensional incompressible turbulent flow within a pump impeller of a torque converter was analyzed numerically at three different speed ratios between the pump and the turbine. The governing equations involving the standard k-ε model in the physical component tensor form were solved with a boundary-fitted coordinate system fixed on the rotating impeller. The computed results were compared with those obtained by the flow visualization and good agreement between them was obtained. Moreover, the results at the three speed ratios were also examined in order to clarify the behavior of secondary flow patterns in detail. The results showed that the structures of secondary flow patterns were influenced by the correlations among the intensities of the Coriolis force and the centrifugal forces caused by the passage curvature in the meridional plane and the blade curvature on the blade-to-blade surface.

抄録全体を表示

An experimental study has been carried out to investigate the melting heat-transfer characteristics of a horizontal ice cylinder immersed in quiescent saline water. The measurements were carried out for saline water of 3.5wt% salinity for ambient temperatures ranging from 1.8 to 19.8°C. The effect of the liquid depth from the saline water surface to the top of the initial ice-cylinder surface on the melting characteristics was examined extensively. It was found that the melting heat transfer at the upper portion of the ice cylinder was influenced markedly by the liquid depth, while that at the lower portion of the ice cylinder was negligibly varied by the liquid depth.

抄録全体を表示

The development of an energy-supply system which is very efficient and does little harm to the environment is desirable. Adsorption heat pumps satisfy these requisites because they use low-level thermal energy as a driving force, and use safe and natural substances ; silica gels or zeolites as adsorbents and water as a refrigerant. A tube plate reactor in which adsorption and desorption take place is proposed here to make the system compact, to enhance heat transfer, and to provide increased heat storage capacity. Fundamental heat-transfer experiments are carried out. And the heat-transfer mechanism is identified through a comparison of the experiments and simulations. The concept of equivalent thermal conductivity is defined, and as a result, it is also identified, which is essential for the thermal design of the heat pumps.

抄録全体を表示

The effect of imposed pulsation on turbulent heat transport in fully developed turbulent pipe flow heated with a constant wall heat flux was investigated. Particular attention was paid to the condition where the pulsating frequency was slightly less than the bursting frequency. In this case, phase-averaged quantities such as the Nusselt. number and the relative correlation between velocity and temperature fluctuations show a substantial phase-lag to the pulsating flow field. Thus, the structural change in the heat transport mechanism takes place only when pulsation is imposed at a frequency smaller than the natural bursting frequency.

抄録全体を表示

Characteristics of an excess enthalpy flame stabilized in an axisymmetric flow impinging on the heat receiver wall (the stagnation surface) in a combustion chamber, and heat fluxes utilized from flames under the conditions of high flow rates near the extinction limits were experimentally examined. The effects of Lewis number of the deficient reactant on the characteristics of combustion were also discussed. (1) The excess enthalpy flames are stabilized near the heat receiver wall due to the effect of heat recirculation for the lean methane-air flames (Le≒1.0) and the lean propane-air flames (Le>1.0), and due to the effects of heat recirculation and the Lewis number for the lean hydrogen-air flames (Le<1.0), even in cases in which a large amount of heat is utilized from the flames. Thus, the heat flux utilized from excess enthalpy flames of methane, propane and hydrogen through the heat receiver wall increases rapidly with an increase in the stagnation velocity gradient. (2) When the flame is established outside the thermal boundary layer formed on the heat receiver wall, heat flux obtained from the flames through the heat receiver wall can be considered as the Convective heat transfer in the stagnation flow in which only the burned gas of high temperature flows. In contrast when the flame is stabilized inside the thermal boundary layer formed in the stagnation flow in which only the burned gas of high temperature flows, the heat flux transferring through the heat receiver wall increases rapidly with an increase in the stagnation velocity gradient, and the characteristics of convective heat transfer to the heat receiver wall are improved. The characteristics of heat transfer to the heat receiver wall are not dependent on the thickness of the thermal boundary layer, but are significantly affected by the flame location.

抄録全体を表示

A measuring system for local heat transfer coefficients on surfaces of blades and a platform of the cascade was constructed. Temperature measurements were carried out from 5 directions to create a complete temperature image of the blade using an infrared camera. The local variation in heat transfer coefficients of the blades was investigated for two inlet boundary layer thicknesses and two turbulent intensities. The accuracy and spatial resolution of the measurements were evaluated.

抄録全体を表示

Shamsundar's method, in which heat exchanger effectiveness of a latent heat storage unit is calculated using a similarity curve representing the relationship between similarity function and frozen fraction, has a drawback that release of sensible heat is always neglected after the frozen fraction at the inlet of the storage unit becomes unity. On the basis of the idea that the release of sensible heat will be taken into account if the heat release fraction is used as an independent variable instead of the frozen fraction, a new method for calculating the heat exchanger effectiveness utilizing the similarity curve of the heat release fraction has been proposed and developed. The comparison between numerical and experimental results on a concentric-circular-tube-type storage unit with n-eicosane as a phase change material shows that the new method is especially advantageous in the solidification process after the inlet frozen fraction exceeds unity. In the similarity curve of the heat release fraction, however, the similarity rule is less applicable than in that of the frozen fraction.

抄録全体を表示

An experimental study was conducted on the evaporation and combustion phenomena of a single n-dodecane drop in ambient gas pressurized and heated beyond the critical point of fuel, that is, supercritical environments produced on a hot wall. The evaporation lifetime in air at the wall temperature below the ignition temperature was found to be shorter than that in nitrogen at the same wall temperature. The end of combustion time in air at ambient pressure below the critical pressure of the fuel was correlated with the evaporation lifetime, and it was maximum in the wall temperature range where the drop showed spheroidal evaporation. As the ambient pressure increased beyond the critical pressure of the fuel, the end of combustion time increased in the wall temperature range around the critical temperature of the fuel, and in other wall temperature ranges, it decreased.

抄録全体を表示

An experimental study was conducted on the evaporation and combustion phenomena of a single fuel drop in ambient gas pressurized and heated beyond the critical point of fuel, that is, supercritical environments on a hot wall. Several kinds of liquid paraffin such as n-heptane, n-decane, n-dodecane and n-tetradecane were used as the test fuel. The evaporation lifetime in air was found to be shorter than that in nitrogen at the same wall temperature. The difference in the evaporation lifetimes in air and nitrogen environments was greater under higher ambient pressure for the same fuel ; for the same ambient pressure, it was greater under a larger ratio of ambient pressure to critical pressure of the fuel. As ambient pressure increased from subcritical to supercritical pressure, the end of combustion time increased in the wall temperature range around the critical temperature of the fuel, whereas in a wall temperature range higher than the critical temperature of the fuel, it decreased.

抄録全体を表示

Characteristics and reaction mechanisms of N₂O formation/destruction in high-temperature pulverized coal combustion are studied experimentally and theoretically. The experimental results show that N₂O is mainly formed during the initial stage of volatile matter combustion, but that the N₂O concentration at the exit of the furnace reaches a low value. The maximum N₂O concentrations in different types of coal depend on the temperature gradient, the content of volatile matter and the flame structure around coal particles. N₂O concentration at the exit increases with increasing air ratio. By numerical analysis of combustion behavior using a single-particle combustion model, it becomes clear that N₂O is mainly produced by NH (formed by the decomposition of NH₃) and N₂ in the initial volatile matter combustion region, after which N₂O is formed by NCO produced by the oxidation of HCN ; however, N₂O is decomposed by H radicals.

抄録全体を表示

Coal-gasified fuel contains ammonia in a hot gas cleaning system in integrated coal gasification combined cycle (IGCC) power generation systems. Ammonia is converted to nitrogen oxides in the combustion process of a gas turbine. Research data on the NO_x formation mechanism in fuel is insufficient, and there still remains a wide unexplored domain. The present research aims at obtaining more fundamental knowledge by means of elementary reaction kinetics. We have obtained a very significant result by comparing the kinetics calculations with combustion experiments. The result is that fuel NO_x formation characteristics is affected considerably by the conditions of secondary-air mixing and the concentration change of methane in fuel.

抄録全体を表示

Flow patterns in a swirl flow combustor, combustion processes and flame stabilities were experimentally investigated with two types of internally mixed twin-fluid atomizers, slit nozzles and multi hole nozzles, and conventional hollow-cone nozzles. The results show that, (1)twin-fluid atomizers can realize high-intensity spray combustion at high-velocity spray flow in comparison with a hollow-cone nozzle, due to intense mixing of spray with air and recirculating hot gas, but at small air-swirl degrees or with a small recirculation zone, flame stability becomes worse with the increase of flame length, (2) spray of slit nozzles has a deflating tendency at ordinary spray angles, but can extend a narrow recirculation zone at large spray angles, which is effective in expansion of the flame stability region. Multi hole nozzles are preferable to maintain a constant spray angle and significant penetration.

抄録全体を表示

The exciplex-based planar fluorescence technique was applied for two-dimensional visualization of a fuel spray in the region close to the nozzle tip. A spray was discharged from a diesel nozzle into pressurized gaseous nitrogen inside the test chamber. The spray was also allowed to evaporate in the high-temperature environment produced by combustion of a homogeneous mixture of methane and air. Photographs of the temporally frozen image of the fuel spray were processed using an image analyzer. The results showed that the high-density portions of liquid fuel appeared almost periodically downstream and that the axial distance between neighboring high-density portions increased with an increase in the downstream distance. It was also observed that the surface wave appeared at a location submillimeters downstream from the nozzle tip.

抄録全体を表示

Exhaust particulates in diesel engines are affected by the fuel properties, especially the aromatic hydrocarbon content and distillation properties. However, the reasons for this fact are not clear. Particulate formation has been reported to start with thermal cracking of the fuel to lower boiling point hydrocarbons, followed by condensation polymerization and production of benzene ring compounds. The formation of particulates takes place via polycyclic aromatic hydrocarbons. This report investigates the thermal cracking and condensation polymerization process of various fuels up to the solid particulate formation temperature zone, using a high-temperature thermal cracking apparatus in nitrogen atmosphere. The following results are obtained : 1 Depending on their molecular structure, aliphatic hydrocarbons start thermal cracking and condensation polymerization at slightly different temperatures, but the formation processes of particulate is almost the same no matter what the carbon bonds. 2 Benzene rings of aromatic hydrocarbons are almost never cracked in the formation of the polycyclic aromatic hydrocarbons.

抄録全体を表示

In a four-valve spark-ignition engine with pentroof combustion chamber, the tumbling flow and its turbulence in compression stroke were measured with a laser Doppler velocimeter in two kinds of compression ratios under the condition of motoring. The main results are as follows. In the latter stage of the compression stroke under the condition of higher compression ratio, the tumbling flow is broken down to several small-size eddies because of narrow top clearance, and the turbulence intensity increases. Then the power spectrum of the turbulence is increased in the higher frequency region at the top dead center of the compression stroke. In the early stage of the compression stroke, the effect of the compression ratio on the power spectrum of the turbulence does not appear due to use of the same intake system.

抄録全体を表示