Numerical simulations are conducted for behaviors of reflected-shock waves interacting with side-wall boundary-layers in shock tubes. A finite-difference procedure is developed as a solver of the thin-layer Navier-Stokes equations. This procedure, which combines the Lax-Wendroff-FCT scheme with the Crank-Nicolson scheme, is applied to initial and boundary conditions in which effects of shock-tube side-wall boundary-layers. are taken account of by use of the Pohlhausen polynomials. Computations are carried out for reflected-shock waves in nitrogen and in argon gas, respectively. Bifurcated reflected-shock waves, observed experimentally in polyatomic gases in shock tubes, are reproduced in simulations for nitrogen gas and compared with semiempirical theory for bifurcated reflected-shock waves. Good agreements between them lead us to conclude that the present procedure can simulate essential parts of interactions between reflected-shock waves and side-wall boundary-layers.

抄録全体を表示

Fully developed gas-solid two-phase flow in a vertical pipe was simulated numerically, and solid concentrations and velocities were predicted. Particle motion in a pipe which has periodic boundaries on the inlet and the outlet was solved, taking into consideration particle-to-particle collisions. To investigate the effects of particle-to-particle collision on the particle flow, simulations were performed. It was found that particle-to-particle collisions have a large effect on the diffusion of particles in a pipe cross section even at a small concentration. The predicted solid concentrations and velocity profiles agreed well with the experimental results.

抄録全体を表示

Experiments were performed on the interaction between the reflected oblique shock wave and the separation shock wave ahead of the forward-facing step through the turbulent boundary layer. The turbulent boundary layer disturbed by the oblique shock reflection recovered exponentially in the downstream direction when the step was absent. The interaction was divided into three types according to the position of the step : (I) the reflection of the oblique shock is independent of the separation shock, (II) the recovery of the boundary layer is insufficient, and the separation shock becomes weak and merges with the reflected shock, (III) the new separation shock ahead of the incident shock plays the role of the reflected shock wave. These type were characterized by the shape factor of the boundary layer.

抄録全体を表示

Wall pressure fluctuations are measured experimentally regarding the pulsatile flow of Bingham plastic fluids in a high Reynolds number region. The wall pressure fluctuation is a very interesting quantity from a biomechanical viewpoint (causes of arterial damage) and from a practical viewpoint (causes of the fluctuating stress acting on a pipe). The experimental results are summarized as follows : (1) The rms (root mean square) of wall pressure fluctuation in the deceleration phase is greater than that in the acceleration phase for both the Newtonian fluid and Bingham plastic fluid. (2) The non-Newtonian property of the fluid increases the difference of the rms of wall pressure fluctuation between the acceleration phase and the deceleration phase. (3) The difference in power spectrum density between the acceleration phase and the deceleration phase is clear in the low-frequency region below 100 Hz.

抄録全体を表示

Magnetic fluids have certain unique properties which are very important in physics as well as in engineering. Recently, it has been reported that ferromagnetic particles form clusters even in a weak magnetic field. No one has analyzed the cluster phenomena in detail. In the present paper, a new statistical equation is derived by considering the cluster phenomena using a transport theorem in statistical mechanics. The basic concept is an extension of the Bolzmann equation for plasma which is applied as a model of the combination of ferromagnetic particles in the collision process and dispersion of cluster by the phase transition from a microscopic viewpoint. By the present method, we can calculate the increment of apparent viscosity under the contribution of cluster and the critical magnetic field where clustering appears.

抄録全体を表示

The fully developed periodic laminar flows of incompressible Newtonian fluid through a pipe of circular cross section which is coiled in a circle have been simulated numerically. The flow patterns are characterized by three parameters: the Womersley number W_o, the amplitude ratio β, and the Dean number D_e. The effect of these parameters on the flow has been studied for the range 15.07≤D_e≤265.49, 2.l9≤W_o≤50.00, and 0.50≤β≤2.00 for the curvature ratio, δ, equal to 0.05. The mannner in which the secondary flows evolve with increasing Womersley number and Dean number is explained. The secondary flow patterns are classified into three groups : the viscous-term-dominated type, the convection-term-dominated type, and the inertia-term-dominated type. It is found that, when the amplitude ratio of the volumetric flow rate, prescribed by a cosine function, is equal to 1.0, 4 - 6 vortices of the secondary flow appear at high Dean number, and that the Lyne-type flow patterns disappear at β≥0.50.

抄録全体を表示

A numerical analysis technique for free surface flow problems by FEM is presented. The VOF (Volume of Fluid) method is employed in order to represent the free surface by Eularian mesh. FEM with the VOF method mentioned here is convenient to simulate the large deformation of the free surface without re-meshing. The convective equation of VOF is solved by a fully implicit scheme, thus simulating shapes and locations of the free surfaces accurately. Numerical calculations were carried out for the breakdown flow of a dam and the mold filling simulation in die casting and these results were examined in comparison with experimental results and calculation results in the references.

抄録全体を表示

Wave characteristics of the swirling flow due to a single swirling vortex layer in a rigid circular pipe were analyzed by the linear theory assuming that the disturbance was small and the flow, except for part of the vortex layer, was irrotational flow, and the influence of the diameter of vortex core, density ratio, and axial flow velocity on the stability of the flow was investigated. It was clarified that the swirling flow became stable due to the restoring effect of the centrifugal force generated by the flow of the free vortex which was situated on the outer side of the disturbed vortex layer, and was destabilized or stabilized according to the equality or the inequality of the virtual mass of fluid on both sides of the vortex layer.

抄録全体を表示

Three components of mean velocity and velocity fluctuation have been measured in a turbulent flow through a square-sectioned 90°bend without a straight downstream duct and near a bend exit. The measured data, such as the mean velocity, the turbulence energy and the correlation of velocity fluctuations, are compared with results obtained in the bend with the downstream tangent, and effects of the downstream tangent on the flow field are discussed.

抄録全体を表示

An artificial lung used as a supplementary means during open heart surgery is one of the artificial organs. At the present time, the artificial lung of bubble oxygenator and membrance oxygenator are in clinical use. Both type of the oxygenators have merits and demerits, respectivelly. For the bubble oxygenator, the destruction and degeneration of the components in blood come particullarly into question. In this work, noise spectra generated in the internal flow of the bubble oxygenator is measured. The noise generated at high flow volume of O₂ attains high sound pressure level, which would be the cause of hemolysis. The bubbles generated in the mixing chamber of this artificial lung are observed by a strobo photographing, then the bubbles are 0.5 to 4 mm in diameter. By comparison between the natural frequency of the bubbles and the frequency of noise spectra, it seems that the noise generated in the bubble oxygenator is caused by the natural motion of bubbles.

抄録全体を表示

To determine the optimum geometries of louvered fins, a parametric study on fin pitch, louver angle and fin thickness was performed with a two-dimensional finite difference code developed by the authors using an overlaid grids method. The numerical results showed the existense of the most effective ratio of fin-pitch/louver-pitch for each louver angle, which controls the thermal wakes formed after louvers. From the discussion about the effect of thermal wakes, a numerical expression that estimated this optimum relation among the fin parameters was proposed. The estimation of the performance of heat exchangers consisted of these optimized louver fins showed that a heat exchanger with smaller louver angle(θ) had superior performance in the range of 20°≤θ≤30°.

抄録全体を表示

The vortex method is one of the methods used to solve high Reynolds number flow problems of incompressible, inviscid fluids. This method has been used for various problems : vortical flows in aerodynamics, separated flow about cylindrical bodies, chaos, stability of the row of vortices in a two-dimensional flow. Mathematical research of this method has also been carried out and the convergence theorem has been established. Nevertheless, some researchers say that this method is awkward because of the appearance of the very large induced velocity and the difference, in calculated results due to the time steps, and so on. To resolve such problems, it is important at least to reduce this method from the Navier-Stokes equation directly. So in the present paper, we aim to formulate the vortex method from the Navier-Stokes equation directly, and to clarify the relation of the earlier proposed vortex methods.

抄録全体を表示

This paper describes the image processing of visualized streak lines. Two-dimensional steady jet flow from the restriction of a spool valve is visualized by the hydrogen bubble method. The photographs of streak lines are taken with a 35 mm camera, and the coordinates of streak lines are fed into a microcomputer using a digitizer. Since streak lines coincide with streamlines for steady flow, a streamline coordinate system is used to conveniently calculate free streamlines of the free jet flow. A flow net, consisting of streamlines and curves orthogonal to the streamlines, is constructed. Finite difference approximation is applied to the basic equations of the streamline coordinate system. The velocity, vorticity and pressure at each node of the flow net are calculated using finite difference equations. To estimate the accuracy, these values are substituted into the equations of continuity and metric coefficient. The calculated results of residuals show that the present method is useful to analyze such a flow with free streamlines.

抄録全体を表示

Decomposing each term in the transport equation of the turbulent kinetic energy, an anisotropic k-ε model, which is valid right up to the wall, is proposed. The deviation of the Reynolds stress tensor from the common isotropic representation is given as a function of nonlinear quadratic terms of mean velocity gradients as well as anisotropic diffusion terms of turbulent kinetic energy. Trial calculations carried out for two-dimensional turbulent channel flows demostrate that this anisotropic k-ε model is able to predict reasonably well the strong anisotropic limiting behavior of the normal Reynolds stresses towards the wall, which cannot be dealt with by the previous isotropic and anlsotropic k- E models.

抄録全体を表示

The anisotropic k-εturbulence model of Myong and Kasagi (1988), which is valid right up to the wall, is applied to the prediction of boundary layer flows with and without pressure gradients. Predictions are made first for flat-plate boundary layer flows with zero pressure gradient, and then for self-preserving sink flow boundary layers with constant acceleration parameters and a nonequilibrium boundary layer flow with a strong adverse pressure gradient. Extensive comparisons are made between the present numerical solutions including the distribution of each Reynolds stress component and the results of previous experiments and direct numerical simulations. It is shown that the model performance as a whole is very good.

抄録全体を表示

The Reynolds-stress model, proposed by Shima and applicable right up to a wall has been partly modified and the performance of this model has been investigated in turbulent channel flow. The mean velocity, the turbulence intensities and the dissipation rate of the turbulence kinetic energy are calculated. It is found that in all respects investigated in this paper the modified version of the Reynolds-stress model is superior or at least equal to the existing Shima model and anisotrokic k-ε model. Next, the proposed model is applied to turbulent Couette flow to conform the universal validity of this model. In this flow, the turbulence has two strong characters. Near the wall the three components of the turbulence intensities differ significantly and are similar to those of the wall turbulence. But the core region exhibits strong indications of the homogneity in turbulence structure. The proposed model describes these properties quantitatively.

抄録全体を表示

Time-averaged velocities of axial and secondary flow components and a turbulent intensity were measured by the hot-wire anemometer in a developing flow through a π bent tube or S-duct composed of two π/2 bent tubes. In the π-bend, the secondary motion is intensified behind the inlet to it and weakened at its outlet. The secondary flow grows rather in the hind bend of the S-duct than in the fore one, though the centrifugal force is in the opposite direction in the two bend. The turbulence intensity is larger near the inner wall than that near the outer wall in the hind bend.

抄録全体を表示

A numerical analysis has been made on three-dimensional developing turbulent flow in a square duct using a modified Reynolds stress equation model and transport equation model for turbulent flux. The calculated results for the velocity field describ a characteristic phenomenon of developing turbulent flow in comparison with the available experiment. The results for the temperature field suggest that the value of the turbulent Prandtl number which was used as a constant in the two-equation model actually changes along the flow direction.

抄録全体を表示

A direct simulation (DS) by Spalart of a three-dimensional turbulent boundary layer created by a rotating free-stream velocity vector offers a standard test case for turbulence models. The flow situation is calculated here using a second-moment closure applicable up to a wall, which was previously shown to reproduce various two-dimensional flows. The numerical solutions are extensively compared with the DS results including the distributions of all the Reynolds stress components. The agreement is generally satisfactory, showing the wide applicability of the turbulence model. The rather close alignment of the Reynolds stresses with the mean shear in the flow suggests a need for a separate test on more strongly three-dimensional boundary layers.

抄録全体を表示

When a simplex finite element method is applied to solve a linear differential equation governing a region enclosed by curved boundaries, the boundaries are obliged to be represented by a polyline. Boundary conditions of the second kind are specified as the derivatives normal to the boundaries, causing numerical error in the results. This paper concerns a finite element method using linear-triangular elements, in which boundary conditions of the second kind are introduced more accurately. When the curved boundaries move or rotate, the resultant error is remarkably improved, and the number of nodes can be decreased.

抄録全体を表示

As the flow through a cascade should satisfy the Kutta condition and blades have generally a finite thickness at their trailing edge, it has been difficult to realize computationally the Kutta condition in the present finite element methods. In this paper, a method to realize the Kutta condition is proposed by making the pressures on the pressure-and the suction-sides of the trailing edges equal, and by expressing the blade and small separation region postulated behind it with spline-fit curves. The resultant slip factor gives almost the same value as that proposed by Wiesner or Stodola, and the calculated pressure distributions around the blade show a good agreement with measurements.

抄録全体を表示

Effect of the geometrical configuration on the switching mechanism in a wall-attachment fluid amplifier was analyzed with the experiment using the enlarged device model and water as a working fluid. The type of switching, switching time, switching flow rate were investigated under the condition with various dimensions of three geometric variables which are the length of the side wall, the splitter distance, and the side wall set-back. Flow mechanism of the separation of the attached jet from the attachment wall and the mechanism of the attaching of the jet to the opposite side wall were discussed by using the flow visualization and the pressure distributions during the switching process.

抄録全体を表示

The problem of free convectional boundary layers over a wedge with uniform suction or injection is numerically studied in this paper. The ordinary equations of transformed boundary layer equation in terms of dimensionless stream function were obtained by means of the difference-differential method, and the solutions subjected to the boundary conditions can be expressed in the form of integral equations. The solutions were obtained solving the integral equations by iterative numerical quadrature. The velocity profiles and the temperature profiles were computed for various values of suction/injection parameter, Pradtl number and wedge angle.

抄録全体を表示

In this paper, analyses of the molten metal behavior with free evaporation interface are presented. The evaporation interface is depressed and deformed by the vapor pressure in addition to the surface tension of the molten metal itself. The analysis of the free evaporation interface behavior of the molten metal was performed as the time-dependent problem, where the validity of the free surface boundary conditions was confirmed by means of the calculated value of the phase-propagating velocity of the interfacial wave. And the characteristic time, derived from the vapor pressure, offers an effective way of calculating the time-independent shape of the free evaporation interface and the stationary secondary flow field induced below that.

抄録全体を表示

Both the aerodynamic- and the noise-characteristics were investigated focussing on the effects of various parameters such as the axial position of propeller fan relative to the duct, the duct width, the shape of the duct inlet and the tip clearance in order to design a lower-noise and higher-efficiency propeller fan. In this paper, the aerodynamic characteristics are discussed. The peak efficiency increases to a level much higher than that of an ordinary ducted fan provided that the location of the propeller fan is beyond the upstream end of the duct. The peak efficiency is highest when half of the propeller fan width is outside the upstream end under the condition that the duct length is about a half of the fan width. Also the peak efficiency increases as the tip clerance becomes smaller.

抄録全体を表示

The effects of various parameters of a propeller fan, such as the axial position of the fan rotor relative to the duct inlet, the duct width, the inlet shape and the tip clearance on the fan noise, have been investigated systematically in order to make clear the optimum conditions for a low-noise fan. In this research the noise characteristics are investigated in relation to the aerodynamic characteristics discussed in the previous paper. It has been clarified that there is a close relation between noise and aerodynamic characteristics, that is, the higher the peak efficiency, the lower the specific sound noise K_SL if the noise level is rated by the L level of a sound level meter. If it is defined by A level, however, the specific sound level K_SA does not show the same relations that K_SL has with the parameters tested. This is due to the difference in the spectrum distribution of the noise.

抄録全体を表示

In the course of the work of newly established JSME Standard S-008 'Performance Conversion Method for Hydraulic Turbines and Pumps', an accurate and relatively simple method of caluculating leakage flow and disk friction power has been developed, and the simple formulae for volumetric efficiency and mechanical efficiency at the best efficiency point have been determined. This paper presents the above method and formulae, and shows their validity comparing with the measured data and the numerical analysis.

抄録全体を表示

Theoretical and experimental study has been made for a newly designed flowmeter. The flowmeter consists of an axial turbine rotor, a d. c. motor and so on. Angular velocity of the motor is controlled to be kept constant. Output signal of the flowmeter is the motor torque which is related to instantaneous flow rate. Theoretical analysis on the output characteristics of the flowmeter has been done under steady flow and pulsating flow with small amplitude of fluctuation. Pulsating quantities are treated as the time average value and the time fluctuating part. The characteristics of the flowmeters testing have been obtained experimentally under the same flow conditions as that in the theoretical analysis. Experimental results showed good agreement with the theory. It is concluded that this flowmeter is appropriate for flow rate measurements of steady and pulsating flow with small amplitude.

抄録全体を表示

Unsteady flow field inside a Savonius rotor at various overlap ratios has been investigated by flow visualization with a smoke-wire method. The experiment is performed for rotors with 4 different overlap ratios (OL=0-0.5) and for various rotor angles and tip-speed ratios. It is found from the experiment at the optimum overlap ratio (OL=0.15) and tip-speed ratio (X=0.9) that the extra power due to the overlap is mainly produced by the pressure recovery effect of the flow through the overlap to the returning blade without flow separation. For the rotors with rather large overlap ratios (OL=0.3-0.5), the separated region grows on both the blades, and it prevents the flow through the overlap. These phenomena, together with the decrease in the drag force acting on the advancing blade, indicate the decrease of the net power at large overlap ratios. This mechanism agrees qualitatively with the measured power performance of the Savonius rotor with overlap.

抄録全体を表示

The effects of the shape of the blade, the tip clearance and the sectional profile of the tip blade on fan noise were investigated experimentally. Analyzing the boundary layer for the suction side of blade, the effects of the seperation point on sound pressure level were clarified. It was found that the most suitable construction for the fan noise reduction was the blade inclined toward the rotative direction, the small tip clearance and the cut-off blade of the tip section in the pressure side.

抄録全体を表示

An experimental study has been made of the critical heat fiux (CHF) of natural convective boiling in vertical uniformly heated tubes. The experiment was performed at a pressure of P=0.1 to 3.1 MPa for the tube diameter of D-1.12 to 18.4 mm, the tube length of L=23 to 960 mm and three kinds of liquids. The effects of the pressure which corresponds to the density ratio pυ/pι ranging from 6.24×10^-4 to 0.236 and of the ratio of L/D=5 to 857 on the CHF, are mainly discussed. The experiment shows that the characteristics of the CHF can be first divided into two regimes at a point where a nondimensional parameter of D/√(σ/g(pι-pυ)) is equal to about 13. In addition, the CHF in D/√(σ/g(pι-pυ))<13 can be further divided into two regimes ; the CHF of L/D < 240 takes place under the condition of slug or annular flows, and then the CHF data can be predicted with an accuracy of ±20% by Eq. (2), and the CHF of L/D=857 may be caused by flooding. On the contrary, the CHF in D/√(σ/g(pι-pυ))<13 occurs under the same conditions as the CHF in pool boiling. Then the CHF data are also correlated by Eq. (4), in which the tube diameter has no effect on the CHF.

抄録全体を表示

From the view point of energy saving and the development of new energy resources, it is important to utilize geothermal resources and waste heat from factories. As one of the energy conversion expanders, there is a radial outflow reaction turbine (that is, Hero's turbine). Performance characteristics of Hero's turbine using subcooled hot water as a working fluid are clarified analytically and experimentally. It is found that : (a) there is an optimum rotational speed at which maximum turbine efficiency can be obtained, (b) Hero's turbine internal efficiency can be expressed by an algebraic equation (7), and (c) nozzle loss amounts to almost 90 percent of the total turbine loss.

抄録全体を表示

The method of numerical analysis is investigated for the maunfacturing processes by means of electron beam heating such as hardening, cutting and welding. The high-energy electrons (10-50keV)impinge upon the surface of the material and diffuse through multiple elastic/nonelastic scattering with atoms. Although the electron collisions with atomic nuclei acn approximately be treated as elastic ones, the collision with the orbital electrons of atoms is a nonelastic one. The fast electrons are decelerated in the course of the atomic excitation or X-ray radiation, transferring their kinetic energy into the lattice system as heat energy. Here, the difference between the heat generating density and the electron density is clarified numerically, as are the penetration depth and reflection ratio of the electron beam, which give good agreement with the referenced data. Furthermore, the difference between the penetration depth of the electrons and that of the heat, which has never before been discussed in detail, is clarified.

抄録全体を表示

The enhancement of heat transfer around a rectangular cylinder in a flat plate boundary layer is very interesting. The side length of the cylinder was 30 mm, and the height of the cylinder was varied at 3, 5 and 10 mm. The velocity of undisturbed flow ranged from 2 to 6 m/s. In this case, the transition from laminar to turbulent flow was achieved by a tripping wire of 2.0 mm in diameter. The local heat transfer coefficients on each face of the cylinder were measured and discussed in relation to the flow characteristics. The heat transfer coefficients around the cylinder of the turbulent flow were 1.5 to 3.0 times that of the laminar flow.

抄録全体を表示

Heat transfer measurements have been made for a condenser consisting of fine tubes (outside diameter : 1 mm) and laterally woven threads (diameter : 0.3 mm). Using steam as a working fluid, tests were performed at atmospheric pressure. Effects of steam velocity, noncondensable gas, and orientation of the condenser were investigated. Since the threads intensify the vapor shear and act as fins, heat transfer of the fine-tube condenser with threads is enhanced markedly compared with that for the case without threads, and the maximum heat transfer coefficient reaches 200 kW/(m²K)(the maximum heat flux is 2 000 kW/m²). The high heat transfer coefficient is ascribed partially to the small characteristic length of the fine tubes as well as the effects of the threads mentioned above.

抄録全体を表示

Numerical computation has been carried out for the laminar forced convective heat transfer in eccentric annuli. The finite difference equivalents of the governing equations of velocity and temperature fields written with a bipolar coordinate system were solved with an iterative procedure. A study has been made for two types of the second kind of thermal boundary conditions. The friction coefficient and average Nusselt number obtained were found to agree with the results previously reported by others. This validates the present computational method. The obtained cross-sectional distributions of velocity and of temperature, and the peripheral distributions of wall temperature and of local Nusselt number were discussed in detail. When the radius ratio is rather close to unity and eccentricity is large, the peripheral distributions of the friction coefficient and the local Nusselt number are remarkably nonuniform in relation to the conspicuous nonuniformity of velocity in the eccentric annulus. The local Nusselt number then takes a negative value in some regions. The wall temperature there is lower than the fluid bulk mean temperature because of a large peripheral nonuniformity of wall temperature. The temperature field in some cases is significantly far from axisymmetry.

抄録全体を表示

Heat transfer characteristics in heated tubes under the periodically reversing flow conditions have been experimentally investigated using the test apparatus that simulates heat exchangers for the actual Stirling engine. It is shown that the heat transfer characteristics under these conditions are greatly affected by the piston phase difference that generates the reversing flow of working fluid, and this phenomena is proper to the heat transfer under the periodically reversing flow which is different from the conventional convective heat transfer in the steady flow. The experimental correlation for the heat transfer coefficient on these conditions is induced by the use of the working gas velocity evaluated from the Schmidt cycle model that is one of the ideal Stirling cycles concerning the influence of the piston phase difference.

抄録全体を表示

This experimental work is carried out to obtain basic information on the melting of frost on a heat exchanger actually used as a heat pump and air conditioner. The dimensions of the heat exchanger are 400L × 44W × 254H mm. First, frosting behavior is investigated to determine the relationship between frost volume and frosting time. It is found that frosting rate is very large until one hour after the start of frosting: after one hour, the frosting rate gradually decreases. Next, defrosting experiments are carried out in the range of 30-400 g of frost. Melting thermal efficiency defined with critical melting time t_c is not affected by frost volume: on the other hand, melting thermal efficiency defined with flow-out time of melting-water t_e is much affected by frost volume. Finally an example of the defrosting experiment is presented, carried out using a compressor working with an actual refrigeration cycle.

抄録全体を表示

An analytical result of the diffusion-controlled melting of a snow layer is presented here when a snow layer was quickly immersed in CaCl₂ aqueous solution. The analytical result predicts the experimental results of melting well. The temperature in the melting system decreases quickly during the melting. The melting completes in a few seconds, which denotes a thermodynamical equilibrium. When the initial temperature Τ_i and the initial porosity of snow ε_i. the constant, the initial concentration C_mi in the solution strongly affects both the decrease in the temperature in the melting system and the melting mass per unit volume Μ. When C_mi, Τ_i and ε_i are the same, the maximum melting mass per unit volume Μ_max in equilibrium is independent of the snow particle diameter. A figure was presented for the relation Μ_max-Τ_i, C_mi, ε_i.

抄録全体を表示

Melting characteristics of phase change material by natural convection heat transfer inside horizontal rectangular capsules are examined experimentally. The capsules are heated isothermally and three kinds of aspect ratios (H/W=3, 1 and 1/3) are provided. Octadecane and ice are used, respectively, as the phase change material. A method of analysis applying the empirical correlations for natural convection heat transfer in a vertical or horizontal enclosure to the melting in the rectangular capsules is presented. The predicted results show good agreement with the experimental data. For the melting of octadecane and ice, it is found that the effect of aspect ratio on melting velocity is not significant for the range of B=1/3 to 3.

抄録全体を表示

New ice heat storage techniques that dynamically strip off ice on the freezer of an ice-water heat storage apparatus have been studied, to increase the freezing speed. However, it is not clarified what ice shape enhances the heat storage capacity of the ice container. Then, rectangular-type ice, small-stone-type ice and snow-type ice are made to measure experimentally the ice filling-up rate in a container. Through the measurement of the rate, it is proved that the filling-up rate of the small-stone-type ice is about 65 % and the highest of all. That of the snow-type ice is about 34 % and the lowest. In the case of snow-type ice, the ice containing brine is occasionally used to decrease the friction factor in the pipe for transportation. Then, the ice filling-up rate and the heat storage capacity of snow-type ice containing brine are measured. As the results, in spite of the fact that the filling-up rate of the ice increases more remarkably than that of the above-mentioned snow-type ice without brine, the heat storage capacity does not increase much. This is because the solidification latent heat of the ice decreases through mixing of the brine. Therefore, the heat storage capacity of the ice is about 1/2 of that of the above-mentioned small-stone-type ice without brine.

抄録全体を表示

Close-contact melting characteristics of phase change material inside horizontal rectangular capsules are examined experimentally. The capsules are heated isothermally and three kinds of aspect ratios (H/W=3, 1 and 1/3) are provided. Octadecane and ice are used, respectively, as the phase change material. A method of analysis applying the Nusselt's theory for film condensation to the close-contact melting heat transfer in the rectangular capsules is presented. The analytical results show good agreement with the experimental data. For the melting of ice, it is found that the effect of natural convection resulting from density inversion of water at 4°C becomes significant for large Stefan numbers.

抄録全体を表示

Many papers have described heat transfer in a regenerator core. We tried to obtain an approximate solution of original partial differential equations dealing with heat transfer between gas and core. The exact solution of the original equations is so complicated that the equations were converted to simultaneous ordinary differential equations by using the finite difference form. We obtained the approximate solution described here and inspected it for accuracy. We then applied it to evaluation of the heat transfer rate of a model core. This method was compared to max. slope method, and the precision of the calculated results was verified.

抄録全体を表示

The present study aims to determine an optimal wind turbine system with battery storage for use in developing countries. In order to utlize wind energy efficiently, we have to make the wind turbine output match with the battery load. If we could control the revolution of the turbine in proportion to wind velocity, we can realize the optimal system. However, the battery load suppresses increase of revolution because the voltage is fixed. Therfore we need to know how to determine the voltage and the capacity of the batttery. The present paper proposes a method of system design based on the computer simulation. We applied it to our Savonius-Darrieus wind turbine (1kW) and confirmed its validity by the field test (4 months).

抄録全体を表示

Experimental investigations on an aerodynamic structure of a laminar boundary layer diffusion flame over a horizontal flat plate have been performed. Effects of gravitational force and velocity gradient of free stream velocity have been discussed. When a fuel (methane) is injected upward, the velocity around a flame zone is locally accelerated, and the velocity gradient at the wall is increased as compared with that of isothermal flow. On the other hand, when a fuel is injected downward, the velocity is retarded, and the velocity gradient at the wall is decreased in not only a downstream region of the leading flame edge but also the upstream region. This reveals that the gravitational force varies the local pressure distribution not only in a downstream region of the leading flame edge but also in the upstream region. It is also elucidated that the velocity gradient of free steam velocity varies the velocity profiles in a boundary layer.

抄録全体を表示

The study on quantitative characterization of flame colors was carried out by using the chromaticity coordinates (x, y) defined by the CIE 1931 standard colorimetric system. Propane/air premixed flames were used in the study. It was found that the coordinates could characterize the colors of flames with various air ratios and contained information about the spectral intensity of CH and C₂ radicals in the flames. These results suggest that flame colors, which had previously been considered qualitative because of their dependence on fallible human perception, can be used to evaluate flame behavior quantitatively. There is a good possibility that the measurement of the colors is applicable to combustion control and monitoring systems.

抄録全体を表示

This study discusses in detail turbulent diffusion flame structures and their similarities by using laboratory-scale turbulent diffusion flames. The characteristics of these flames vary with working conditions, burner sizes, type of fuel, oxidant, etc.. The time scale and PDF of temperature fluctuations were obtained by laser Rayleigh scattering spectroscopy (LRS) which had no direct interference with the combustion field during measurement. A He-Ne laser LDV system was also adopted to measure the velocity fluctuations. The turbulent diffusion flame structure could be classified into four regions by the time scale characteristics of temperature fluctuations. From the results of cross correlations between velocity and temperature fluctuations, it uas ascertained that the turbulent energy flux was not simply proportional to the gradient of the time mean temperature profile in the combustion region where the time scale of temperature fluctuation attained almost the maximum value

抄録全体を表示

This study aims to develop one of the most accurate CARS simulation codes of water vapor and acetylene. The computations include detailed considerations of perturbation phenomena and high energy levels. The water vapor simulation computes Fermi, Darling-Dennison, and Coriolis type perturbations, and the acetylene simulation includes Fermi and ιtype effects. These results are compared with the experimental observations in prior studies, and successful improvements are shown by these simulation codes.

抄録全体を表示

The available energy of a combustion system with energy recirculation between products and reactants is discussed extensively, and it is shown that there exists a theoretical maximum of the available energy of the combustion system corresponding to the adiabatic limit flame temperature of the combustible mixture. Since this limit temperature is an inherent temperature of a combustible mixture based on the thermochemical properties of the constituents, the maximum available energy is also an inherent value of the combustible mixture. It is also shown in this paper that a combustion system with energy recirculation has the potential to realize a higher thermodynamic efficiency from the point of view of generating power even when the allowable maximum temperature in the combustion system is lower than the theoretical temperature of combustion of the reactant mixture of the system.

抄録全体を表示

Numerical simulations on the transient sprays were made based on the Eulerian gas and Lagrangian drop formulations incorporated with the chemical reactions of gases and soot. Computed profiles of species and soot concentration, gas temperature, and local heat release rate were presented to understand the overall and internal structure of the spray ignition and combustion. Some of the computed results were compared with the experiments conducted using a rapid compression machine. The results indicated are as follows. (1) The computations predict the ignition delay, the transient configuration of the spray flame and drop penetration distance reasonably well. (2) Ignition is initiated by a hot spot at the off-axis region followed by the rapid spread of the combustion region which induces an intense heat release rate just after the ignition. (3) The drops evaporate in the early stage of the spray combustion. A flame front is formed downstream from the nozzle exit. Near the flame front, the local heat release rate is very intense where premixed combustion occurs. Gaseous turbulent diffusion combustion is dominant in the downstream.

抄録全体を表示