We investigate the mechanism of drag reduction of a turbulent flow in a channel by wall oscillation. DNS is carried out in which one wall of a two-dimensional channel is stationary but the other is in sinusoidal oscillation normal to the mean flow. The most effective oscillation period is adopted and about 35% reduction in drag is confirmed. For this case, generation terms of quasi-streamwise vorticity are examined from various points of view. It is shown that quasi-streamwise vortices which play a key role in sustaining wall turbulence are most strongly affected in the buffer layer. Vorticity production rate fluctuates with the phase of wall oscillation and is enhanced at some definite favorable phase but suppressed as a whole, in one pitch of oscillation. The stretching effect as well as tilting effect is important howevers, the latter affects mainly weak chaotic vortices and hence the role of the former which modulates coherent quasi-streamwise vortices is more substantial. The suppression of vorticity generation by stretching is due to the decrease of inclination angle of a vortex to a wall.

抄録全体を表示

A suboptimal control theory is adopted to maximize the performance of an active turbulence control in the fully developed channel flow. The control input is a virtual body force field, which damps the spanwise velocity fluctuations within a thin layer adjacent to the wall. The method used is to minimize a cost functional, which is constructed so as to represent a balance of the square body force strength and the square spanwise velocity gradient , the latter represents the dominant compo-nent of either the streamwise vorticity or the strain rate. From a series of direct numerical simulations, it is found that the suboptimal control achieves a turbulent drag reduction as large as that obtained by an unoptimized active control, but the work required for control is to be estimated much smaller.

抄録全体を表示

An experiment study of a turbulent boundary layer that interacts with controlled longitudinal vortex arrays is presented. A row of vortices is produced by thin airfoils placed side by side in free stream. Five cases have been considered in which the spanwise period of vortices and their height are varied. The results show a very strong effect of the spanwise period as well as a notable effect of the height at which vortices are positioned. Namely, the spanwise period controls the vortex path and the vortices interaction with each other. A small period prevents vortices from interacting with the turbulent bounday layer but it also determines an effective vortex/vortex interaction, while a lager period is responsible for the vortices penetrating the inner layer and rebounding near wall and thus strongly interacting with the boundary layer, however, the vortex/vortex interaction is very weak. The vortex path affects the vortex strength : a high path and a small period lead to an abrupt decrement of vortex strength compared to the case when the path is low and the period is large. From the vorticity budget point of view, the small spanwise period leads to high variations of the vorticity flux. The large spanwise period, on the other hand, leads to a steadily decreasing vorticity flux except for a slight variation that is produced by the effect of the vortices axis changing at the rebounding points.

抄録全体を表示

The characteristics of low-speed streaks in turbulent channel flow of a drag-reducing surfactant solution are studied by employing the hydrogen-bubble flow visualization method. The mean velocity profiles and the distributions of streamwise turbulence intensity are also obtained. It is shown that at large values of the drag reduction ratio DR, the profile of the streamwise turbulence intensity has a plateau region around its peak. In the flow of drag-reducing surfactant solution, the mean spanwise streak spacing reaches a minimum at some distance from the wall. In the flow of drag-reducing high density (1 000 ppm) surfactant solution, low-speed streaks appear intermittently at a Reynolds number smaller than those for the 'hump' of the f-Re curve (f : friction factor) . The mean spanwise streak spacing very near the wall increases as DR increases. It is speculated that the nondimensional wall-normal distance y+ of the center of streamwise vortices is larger for the flow of large DR compared with that for Newtonian fluid flow.

抄録全体を表示

The effect of injection, applied through a slit, on turbulent channel flow has been quantified by measuring wall static pressure, mean velocity profile and turbulence intensity downstream of the slit. Mean flow and turbulent characteristics with injection are controlled by injection flow rate, indepen-dent of the injection velocity. Kfirmfin constant in the universal law of the wall is little affected and maintained constant value. In the near-wall region, velocity profile and turbulence intensity recover relatively quickly from the change in boundary condition except close to the slit. Flow in the core region is significantly perturbed by the injection, and turbulent intensity increases with increasing injection flow rate.

抄録全体を表示

The response of a separating flow in a two-dimensional diffuser to an unsteady disturbance is investigated. The disturbance is generated by the turbulent wake of a circular cylinder moving in front of the entrance of the diffuser. Thus the disturbance makes an angle to the main-flow direction. The pressure recovery coefficient, time-mean velocity distribution and phase-averaged velocity distribution are obtained for several disturbances with different frequencies and different angles. Reynolds number based on the velocity of the main flow at the inlet of the diffuser and its height is 5.2 x 104. The pressure recovery coefficient is found to be modified by the disturbance, depending on the frequency of the disturbance. The local deceleration and acceleration are observed near the wall of the inlet of the diffuser due to the disturbance. Thus the disturbance generated by the wake acts as an oscillator to induce a velocity fluctuation at a specific area.

抄録全体を表示

The effects of periodic perturbation on turbulent separated flow in a two-dimensional asym-metric diffuser were investigated. Measurements of velocity fields were undertaken using a LDV, and data reduction was conducted on a time-averaged as well as ensemble-averaged manner that made the mean-periodic-turbulent decomposition possible. Effects of imposed perturbation on the velocity field strongly depend on the frequency of the perturbation, and the optimum frequency ranges between 0.01 ≤ St ≤ 0.04, with St being the perturbation frequency normalized by the channel height at the diffuser inlet and the maximum inlet velocity. At the optimum frequency, the effect of perturbation appears as an organized velocity fluctuation along a separated shear flow, resulting in the enhancement of momentum transport, which is the primary effect of imposed perturbation. An increase in the turbulent Reynolds stress component in the recirculation region is also observed, indicating the secondary effect of the organized fluid motion due to periodic perturbation.

抄録全体を表示

The "lock-in" phenomenon has been investigated experimentally by vibrating cylindrical bluff body surfaces locally using piezoelectric actuators. The maximum amplitude of oscillation is of the order of 20 pm (nearly 1.0 X 10-3 times the body thickness), and the excitation frequency covers below and above a natural vortex shedding frequency from the body. The Reynolds number renges from 5.25 X 103 to 2.31 x 104. Substantial entrainment of vortex shedding frequency has been observed by exciting the upstream surface of sidewalls, where a short laminar separation bubble exists. Vortex shedding becomes more periodical and the total pressure in the wake becomes high by the local excitation of body surfaces. Possibilities for application to the active control of vibration and noise due to vortex shedding, as well as to drag reduction are discussed.

抄録全体を表示

Experimental investigation of separation-reattachment flow control after a forward facing step is performed. In order to control separation-reattachment flow, the surface near the forward facing step is manipulated by piezo actuators set at 3 spanwise locations. It is shown that the spanwise phase of separation-reattachment flow can be controlled without increasing the average turbulence level of the flow by minute movement of the actuators. The spanwise phase can be controlled when the manipulating frequency is between 600 Hz and 1200 Hz. The controllable frequency extends to lower values when the spanwise phase of the actuator movement is uniform, while it extends to higher values for alternate phase manipulation.

抄録全体を表示

The aim of the present study is to clarify the applicability of vortex generator jets (VGJs) in the control of flow separating in a conical diffuser. Systematic experiments were made to investigate longitudinal streamwise vortices induced by eight jets issued from the wall of a circular pipe having 125 mm dia., where VGJs of 6 mm dia. pitched at 30 deg and skewed at 90 deg. The distributions of wall friction show that turbulent mixing in the boundary layer is enhanced by corotating type VGJs. By using a typical VGJ which was installed at the inlet of a 14 deg conical diffuser with an area ratio of 2.56, the separation in the diffuser was almost suppressed, and the pressure recovery coefficient increased from 0.63 to 0.68 at a velocity ratio Vj/U of 1.9. To quantify the structure of flow separating in a diffuser, attempts were made to measure the fraction of backflow with an X-type hot-wire velocimeter.

抄録全体を表示

We describe the effects of a delta tab attached to the exit of an injection hole for passive control of a wall jet. Streamwise and lateral injections of the wall jet into a two dimensional turbulent boundary layer are made. The mass flux ratio of the wall jet to the main stream is set at 0.5 and 1.2. An encapsulated temperature^sensitive liquid crystal is used to measure the temperature distribution on the surface. Timeaveraged velocity is measured using a conventional hot-wire anemometer. Upstream tab with streamwise injection prevents the primary stream from convoluting in the downstream region. Upstream tab with lateral injection increases the width of the^high temperature region in the spanwise direction.

抄録全体を表示

The effect of a streamwise vortex on the development of a wall jet has been investigated experimentally. A streamwise vortex created with a half delta wing mounted on the wall, which has the same height as inner layer thickness, is introduced into a self-preserving wall jet. The mean flow field and its deformation due to secondary flow induced by the vortex are reported in detail. Evolution of the streamwise vortex itself, decay of vorticity, and path and growth of vortex radius are measured and compared with corresponding results obtained in turbulent boundary layers. An increased streamwise momentum flux due to the vortex shows that control with a vortex generator, which is a well known technique for boundary layers, is successful for control of the wall jet.

抄録全体を表示

The diffusion mechanism of an excited rectangular jet was investigated by comparing the measured velocity profiles and the suggested vortical structures. Experiments were carried out using an air jet from a sharp-edged rectangular orifice of aspect ratio 4. The jet was excited in the interaction mode, in which the stable interaction of vortices occurred. The results reveal that the jet diffusion mechanism is closely related to the vortical structures in the rectangular jet : jet width in the major axis plane decreases due to inward flow induced by vortices bending downstream, and that in the minor axis plane increases significantly due to outward flow induced by stretched vortices and split small vortex rings.

抄録全体を表示

Some properties of a pulsating round jet issuing into a uniform cross flow are studied experimen-tally with the aim of developing a device for active control flow separation. The jet is produced by a zero-mass-flow actuator which is a round orifice connected to a speaker driven by a periodic square-wave train. The jet consists of a train of vortex rings, whose pattern can be classified into three types, depending on the circulation of the vortex ring. The trajectory of the pulsating jet is found to collapse into a single curve if it is normalized by the frequency of the square wave, the velocity amplitude at the exit of the orifice, and the cross-flow velocity. The effect of the suction phase of the zero-mass-flow actuator vanishes within one diameter from the orifice.

抄録全体を表示

The objective of the present study is to develop supersonic mixing enhancement techniques for future scramjet engines. Achievement of the necessary mixing enhancement by means of turbulence control is a challenge because compressibility can strongly hinder turbulent motions at supersonic speeds. Here, we propose the manipulation of mixing layers into multiple shear regions in order to alleviate the compressibility effect. To examine this proposal, the stability of compressible mixing layers with combined-hyperbolic-tangent velocity profile is investigated through an inviscid spatial stability analysis. The combined velocity profile has two shear layers : one being in the faster stream and the other in the slower stream. The former is called shear 1 and the latter shear 2. Amplified disturbances are found for each of shear 1 and shear 2. For these disturbances the eigenfunctions and contour plots of the pressure and vorticity disturbances are demonstrated together with their growth characteristics. These results indicate a high possibility that the supersonic mixing can be strongly enhanced by the proposed manipulation. Effects of the convective Mach number on the normalized maximum growth rates are also clarified for the combined-hyperbolic-tangent profile.

抄録全体を表示

We propose unstable streamwise vortices for use in turbulence control, in particular, the super-sonic mixing enhancement for future Scramjet engines. Although their streamwise vorticity is axisymmetric, it exists over an annular region only. Thus the vortex may undergo "inflectional instability" like low-speed mixing layers. Here, we studied the instability characteristics of the proposed vortex placed in a uniform supersonic flow at a Mach number of 2.4 through conducting numerical simulations on the basis of 2D compressible N-S equations and the inviscid stability analysis for the corresponding incompressible flow. Demonstrating the behavior of violent inflectional instability at the linear and nonlinear stages, we discuss the proposed streamwise vortices as a powerful means of mixing enhancement.

抄録全体を表示

The active noise control of discrete tones generated in a uniform jet flow on a two-dimensional wing was investigated. Discrete tone noise is generated by a self-excited feedback loop formed by the acoustic field and the unstable boundary layer. In this work, we conducted an active control experiment using a flap driven by piezoceramic levers, which can vibrate with a phase delayed from the velocity fluctuation signal on the suction side. When the flap motion lags the pressure fluctuation at the trailing edge with a phase angle of 180 degrees, it was found that the discrete tone noise was reduced by a maximum of about 7 dB, the flow fluctuation intensities in the boundary layer on the suction side were reduced by about half, and the correlation area of the flow fluctuation at the trailing edge decreased. This confirms that discrete tone generation is caused by the feedback loop and that the discrete tone generation is actively reduced by trailing edge control.

抄録全体を表示

Unsteady interblade pressure distributions under rotating cavitation and cavitation surge of an inducer were measured using an "AND" circuit which combines the trigger pulses indicating the blade location and the phase of unsteady inlet pressure. This measurement enabled observation of the propagation process of rotating cavitation. In addition, unsteady blade and shaft loading under rotating cavitation can be evaluated and compared with those caused by surge mode oscillations. Although rotating cavitation has smaller inlet pressure fluctuation, it was found to have larger blade load fluctuation than the surge mode oscillation, with its amplitude comparable to a steady load.

抄録全体を表示

The characteristics of cavity flow around a square cylinder were experimentally investigated for various angles of attack (α=0°-45°) and cavitation numbers (δ_υ=0.9-3.9). Measurements of fluid dynamic forces and surface pressure were made and the cavity configuration was recorded with a high-speed video camera. It was clarified that variations of the flow pattern with cavity development depend on the behavior of the separated shear layer. In a perfectly separated flow region (α ≤ 12.5°), fluctuating fluid dynamic forces increase in a specific range of cavitation number. On the other hand, in the reattached region (15°≤ %alpha: ≤45°), the fluctuations reduce monotonically as the cavitation number decreases. At angles of attack α=10° and 12.5° near the reattached flow region, marked increases of the fluctuating forces are caused by the drastic change in the wake structure.

抄録全体を表示

A nonsynchronous vibration caused by a backward-traveling mode of rotating cavitation which had been predicted by linear analyses was observed experimentally. This nonsynchronous vibration was identified using three different methods : analyses of the phase difference against the angular separation of the inlet pressure fluctuation taps, observation of by-product components of a nonlinear interaction of the rotating cavitation with the blade passing component and an observation of cavity length oscillations using high speed film. Experimental results were compared with these analyses. Comparison that the experimental results agreed more closely with those of the analysis without cavity delay than that with cavity delay.

抄録全体を表示

Preconditioned iterative methods in an implicit unstructured grid solver of compressible flows are compared. The three-dimensional Euler and Navier-Stokes equations are discretized by the edge-based finite volume method. The artificial dissipation scheme for the inviscid numerical fluxes is developed for the unstructured grid from Jameson's scheme for the structured grid. The following methods are compared to the supersonic and transonic inviscid flows and subsonic viscous laminar and turbulent flows : the central difference methods with the scalar and matrix dissipation scheme and Roe's upwind method with the MUSCL sheme for the inviscid fluxes ; the scalar and matrix dissipation method for the left-hand side operator in the implicit scheme ; the ILU, DILU and SGS factorizations for the preconditioner ; and GMRES and Bi-CGSTAB methods for the linear system solver. The results show that a combination of the matrix dissipation in the inviscid fluxes and the scalar dissipation in the left-hand side works better than the other shemes. In the preconditioned iterative methods, a combination of DILU or SGS and the Bi-CGSTAB is recommended for the aspect of required memory.

抄録全体を表示

A numerical procedure was derived to perform direct simulations of compressible homogeneous sheared turbulence. In this type of direct simulations, the pseudo-spectral method has often been used with the grid-remeshing procedure. However, this approach causes critical interpolation errors of the aliasing type. In this study, therefore, the grid was fixed and the finite-difference scheme was used in the shear-normal direction. As a result, precise data for compressible homogeneous sheared turbulence were obtained which are useful in analyzing the effect of compressibility on turbulence.

抄録全体を表示

The geometry of flow patterns in numerically simulated compressible isotropic turbulent flows for high r. m. s. Mach numbers was studied using three-dimensional critical point theory. The solution trajectories for three first-order linear differential equations are used to classify the elementary three-dimensional flow patterns defined by instantaneous streamlines. Fluid motions characterized by high rates of kinetic energy dissipation and/or high enstrophy are of particular interest. It is found that motions corresponding to high rates of dissipation are characterized by a 3-D rate-of-strain topology which is of the saddle-saddle-unstable-node type, similar to the compressible mixing layer. Fluid motions corresponding to a high rate of dilatation dissipation are characterized by a topology of the node-node-node type in particular. The influences of Mach number on the geometry of flow patterns are described.

抄録全体を表示

Numerical simulation of the turbulent separated and reattached flow over a blunt flat plate has been carried out using the large eddy simulation (LES) technique. The dynamic Smagorinsky model (DSM) is used for modeling the subgrid-scale (SGS) stress and the Smagorinsky constant is calculated as a function of space and time. Three Reynolds numbers, 50 000, 11 600, and 1 000, are selected for calculations. The results for the former two cases are compared with experimental data and that for the latter with direct numerical simulation (DNS) data. The present results show good agreement with those data, and it is found that LES using DSM is applicable to engineering prediction for the complex flow accompanying separation and reattachment over a wide range of Reynolds number.

抄録全体を表示

Unsteady incompressible viscous flows generated by nutating motion and rotating motion of a circular cylinder are analyzed numerically by the discrete vortex method with viscous diffusion model proposed by Ohgami and Akamatsu. This method is applied to the rotational coordinates. The calculation is executed for Reynolds numbers Re=50, 1 000 and radii of orbit R=3a, 0.9a, 0.32a (d - radius of the circular cylinder). Stream lines show that the rotating cylinder imparts to the fluid stronger motion than the nutating cylinder. The calculated force indicates that the hydrodynamic force caused by the fluid viscosity is directed toward the center of the orbit.

抄録全体を表示

The resistive force theory (RFT) is a conventional method for analyzing the motion of a microorganism swimming with a flagellum. This theory has been used in many analyses ; however, it has some flaws such as that hydrodynamic interaction between the cell body and the flagellum is ignored and the interaction between a portion of the flagellum and the remains is also ignored. In order to estimate these effects, the boundary element method (BEM) is applied to investigate the Stokes flow around a microorganism swimming with a rotating helical flagellum. First, force and torque of prolate spheroids and of circular cylinders are calculated to check the accuracy of the BEM. Fairly good agreement is obtained between the present results and the theoretical ones. Then, the mean propulsive velocity of a microorganism is calculated. The velocity agrees with that obtained by using the slender body theory (SBT) . The interaction between the cell body and the flagellum is small ; therefore, the RFT can be used to evaluate the force and the torque exerted on the flagellum. On the other hand, the interaction among the portions of the flagellum is quite large, and this causes differences in results between the BEM and the RFT. The results of the RFT are improved by adjusting resistance coefficients to include this interaction effect.

抄録全体を表示

The applicability of a cellular automaton technique to engineering problems will be examined by dealing with mixing in multicomponent flow. The quality and accumulation rate of the product depend on the mixing of raw materials basically. After thoroughly understanding the mixing process, we optimize the geometrical allocation of such elements as a propeller with casing, a chimney and a product output nozzle in a reactor. Usually mixing is formulated by partial differential equations of conservation laws and empirical formulae, and solved numerically by the finite difference technique. In order to evaluate the fine structure of time-dependent interfacial behavior in multicomponent flow, the cellular automaton technique is used, since this has an advantage of describing the pattern formation in detail. It will be demonstrated in the present paper that mixing of two-component immiscible flow is reasonably simulated mesoscopically.

抄録全体を表示

In a previous paper, we investigated the natural convection in thermoelectrically conducting fluids in a cubic cavity under a magnetic field applied in the gravitational direction, and complicated fluid phenomena were clarified. Particularly in the case of mercury (Pr =0.025, Gr =3.75 x 107), the numerical results agreed well with the observed experimental results obtained using a thermosen-sitive liquid crystal sheet. In the present paper, we aim to explain the fundamental behavior of thermoelectrically conducting fluids, and numerically determine the natural convection by consider-ing a different direction for the applied magnetic field. The direction of the applied magnetic field is chosen as either the +x, -y or +z direction (see Fig. 1). The convective inhibitory effect of the Lorentz force increases in the order of the +x,-y and + 2-directions.

抄録全体を表示

Fundamental properties of a shock compaction technique of applying converged underwater shock waves have been investigated for rare earth magnet powders. The rare earth magnet made from Fe-Nd-B powder is being developed to have a high magnet performance. Usually this powder is consolidated using static high pressure and bond materials. It is very difficult to obtain over 80% magnet purity and, thus, it is difficult to achieve maximum performance. To solve this problem, shock compaction techniques involving use of explosives were tested and the results are presented here. The equation of state (EOS) of Fe-Nd-B powders is first verified by a quasi -static experiment. The applicability of the EOS is confirmed by a new shock loading test which is proposed in this paper. Both numerical simulation and experiments on the consolidation of the magnet poweders are carried out. A fine-bonded bulk with of over 96% is obtained.

抄録全体を表示

We theoretically and experimentally investigated the relation between sound source and sound intensity field taking real effects into consideration. Our purpose is to obtain the accuracy of the sound intensity technique when applied to aerodynamic acoustics noise such as jet screech. Shock-vortex interaction model was used to predict the sound intensity field. It was tested and compared with experimental measurements. Experimental results showed that the resultant vectors of sound intensity almost focus on a point corresponding to the location of the shock cell. However, the location sometimes deviated from the jet flow field. Furthermore, it was difficult to locate the source of a strong spectrum component. The deviation was considered to be caused by the spread of the source region and the fluctuation of the strong spectrum component was considered to be caused by unsteadiness of jet screech. We clarified the local deviation by the calculation of far field sound intensity using near field sound intensity and the directivity of distributed screech sources.

抄録全体を表示

A photogrammetric measurement method for two-dimensional small-amplitude waves were proposed where a diffuse reflection spot is used as an index point. An equation used to obtain the still water depth was introduced. This equation was confirmed experimentally by using a lser displace-ment sensor which is equivalent to a camera-index-point system in principle. To confirm the applicability of this method to waves form measurement, numerical simulations of measurement by this method were carried out for sinusoidal waves and a composed wave. The results of these simulations show that the small-amplitude waves can be measured with sufficient accuracy when the water surface inclination is small.

抄録全体を表示

This paper describes a direct method for studying vibration of piping systems for reciprocating compressors, based on the requirements given by API-618 Design Approach-3. The reliability of this direct method is examined by some experiments with a test piping system using pressure pulsation and nodal vibration measurements. Overall pressure pulsation amplitude and its frequency com-ponents at each measurement point agree well with the results from numerical analysis. Close evaluation of "pipe-support' stiffness" and boundary conditions are necessary to perform dynamic analysis of the piping system in relation to the pressure pulsation generated by the reciprocating compressor ; they affect the accuracy of the final estimation of nodal displacement distribution of the piping system.

抄録全体を表示

Transonic flows in turbine stages are complex and unsteady. Unsteady phenomena in shock-wave reflection, vortices and boundary layer separation affect the reliability and efficiency of turbines, and the influence of flow through the stage on subsequent stationary blades is large. Not only the analysis of the calculated results but also experimental studies are needed to clarify such unsteady phenomena in the turbine stage. In this study, a small-size turbine was improved by installing stationary blades at the stage subsequent to that which is supersonic. It was investigated whether there are influences of the flow fluctuation caused by rotation of blades on the flow in the channel of subsequent stationary blades behind the rotating cascade, and if there exists generation of vortices at the trailing edge. The waves due to pressure fluctuation corresponding to the period of blade rotation blade were observed on both sides of the trailing edge. Furthermore, it was found that this amplitude of fluctuation on the pressure side was about twice as large as that on the suction side. The mixing of the flows of pressure and suction side makes vortices behind the trailing edge large with Strouhal Number 0.2-0.3.

抄録全体を表示

In fundamental experiments on a sand collector with an injection port at the collecting mouth, the fluidized region was visualized and the effects of particle shape and size on the fluidized character-istics were discussed using spherical glass beads and volcanic ash as the collected particles. As a result, it was concluded that the fluidized volume for both types of particles takes a maximum value at a diameter of about 2001um and the values for both types of particles are almost the same irrespective of the particle diameter. As the diameter becomes smaller than 200 um, the fluidized volume decreases since the contact force among the particles in the particle bed increases. On the other hand, in the case of diameters larger than 200 um, the aspects of fluidization are different for the two types of particles because of increase in the gravity force acting upon the particles.

抄録全体を表示

Labyrinth seal used in rotary machine or gas turbine has normally large amount of leakage which impairs the improvement of performance. In this research work presented hereunder, the seal performance was improved by combining labyrinth seal and brush on the rotor together. The test results showed that more than 50% of improvement of seal performance against the conventional straight type of the labyrinth was confirmed with this arrangement in the stationary state. In addition, very interesting tendencies were also found : The leakage decreased as rotation speed increased and stopped at a certain circumferencial velocity in the case that the brush was provided right behind labyrinth contraction fin (down stream) and rotor was rotated. On the other hand, leakage was increased as the circumferencial velocity increased in the case that the brush was provided in front of labyrinth contraction fin (upper stream).

抄録全体を表示

Critical heat flux during pool boiling on a wire or plate vertical heater has been measured employing water and refrigerant R113. The experiment was carried out using wires 0.5 to 2 mm in diameter and plates 5,7,and 30 mm in width. The height of the heater ranges from 20 to 300 mm and the pressurer were 1 and 2 bar for water and 1, 2, 3, and 4 bar for R113. The results show that in the cases of both wires and the plate 5 and 7 mm in width, a large coalesced bubble entirely surrounds the vertical heater and rises surrounding it, while for the plate with w =30 mm, a large bubble cannot surround and rises along its surface. The CHF characteristic can be divided into two regimes depending on the flow condition when CHF takes place. Correlations are proposed for the CHF of wires and the plates with w =5 and 7 mm, yielding good agreement with the CHF data. The CHF for the plate with w =30 mm has a similar tendency to that of a one side headed surface and can be predicted reasonably well by using the existing correlation for a one side heated surface.

抄録全体を表示

The stability of a very thin film at the initial stage of deposition was investigated using perturbation theory. It is assumed that the critical wavelength of perturbation on a continuous thin film corresponds to the diameter of nuclei and that the surface tension of a thin film changes with the film thickness. With these assumptions, it can be shown that the diameter of a nucleus increases linearly with the deposition time. It is interesting to note that the resulting equation is similar to the well-known equation of contact angle. To confirm the result of the analysis, a vacuum deposition experiment was conducted. Copper nuclei on the substrate were observed using a transmission electron microscope and diameters of island unclei were measured. The results clearly show that the effective diameter of nuclei increases linearly with deposition time. The good agreement between analysis and the experiment indicates the validity of this analysis.

抄録全体を表示

Thin-layer electrodepositions at a single trench on a substrate are investigated using a Monte Carlo simulation. The Monte Carlo simulation method is obtained by extending the simulation method of the diffusion-limited aggregation (DLA) fractal. The method is applied to the coverage problem of a small trench on a substrate. The transport phenomena (diffusion and migration) and the surface reaction are taken into account as the Brownian motion with drift and the sticking probability. It is shown that the transport of metal ions and the surface reaction have an important effect on the morphology of the electrodeposits. It is found that a well-covered electrodeposit is formed when the sticking probability is low.

抄録全体を表示

Experimental and analytical studies of ice formation associated with heat conductor plates located around an isothermally cooled cylinder in crossflow are presented. The heat conductor plates are circular-fins having some clearance between a cooled cylinder and fins in the present study. The analytical results, assuming a quasi-steady-state condition, agree well with experimental data. It is proposed that the onset conditions, which describe whether the ice volume under a steady-state condition is greater or less than that without heat conductor plates, are defined by dimensionless radius, temperature and Reynolds number.

抄録全体を表示

This paper deals with the transient heat storage characteristics of high-density polyethylene pellets used as a latent heat storage material by the condensed heat from the ethylene glycol vapor. The polyethylene pellets are fabricated by the shape-stabilization treatment in which the molten pellets maintain the same shape as the solidified ones. The polyethylene pellets packed in the cylindrical heat storage vessel are melted by the condensed heat from the flowing vapor. The effects of vapor temperature, flow rate and the amount of the pellets on the latent heat storage characteris-tics are investigated experimentally. Moreover, the vapor movement among the packed pellets is enhanced by using the partition plate in the pellets layer. The nondimensional correlation equations of the completion time of heat storage are defined in terms of various nondimensional parameters.

抄録全体を表示

Verification of the applicability of the linearized thermoacoustic theory proposed by Swift, J. Acoust. Soc. Am., 84-4 (1988) , 1145-1180, to an acoustic--resonance tube refrigerator and discussions on the advanced linearized theory were conducted through a comparison with the measured tempera-ture distribution along the stack in a simulated acoustic-resonance tube. The measured temperature in all cases of various stack configurations such as stack plate-spacing and length showed an almost linear distribution along the stack, while Swift's linearized theory gave a curved distribution with relatively large deviation from experimental data. Eddy diffusivity and/or steady streaming effects excited by sound waves were taken into account in the linearized model, and the effects of these terms on the prediction of temperature distribution were examined. The agreement between the theory and the experiment was markedly improved by the introduction of steady streaming. This provided a guideline for the construction of an advanced linearized theory.

抄録全体を表示

The measurement results of spectral insolation are ranked every 100 W/m2 based on global insolation. It is clarified that the spectra for total radiation in each ranking are similar to each other. Although the peak value of the spectra can be found at the wavelength of 0.48 p.m, on a clear day, energy of longer wavelength than the peak in the spectra is more attenuated than that on a cloudy day, so that the slope of the spectra is steep, and the absorption effect in the atmosphere is stronger than the other effect. Normal direct radiation is related to the rank amount of direct insolation or the altitude of the sun, and thus the peak value of wavelength on the spectra shifts from 0.48 j.tm to 0.76 p.m. With respect to the total radiation ranked, the spectra normalized by its peak value gather each other, and ratios of the spectra are converged upon unity with a deviation ±20 percent for all the rank.

抄録全体を表示

In the numerical analysis of diffusion flames, multicomponent diffusion should be correctly evaluated, because various species with different diffusivities are generated in the process of combus-tion. In the evaluation of multicomponent diffusion, effective binary diffusivity is often adopted instead of exact multicomponent diffusion coefficient. For diffusion flames of H2 or H2 diluted with N2, the accuracy of calculation by effective binary diffusivity is investigated by comparison with that of exact multicomponent diffusion. In multicomponent system, diffusion of a species is generally driven not only by its own concentration gradient but also by that of the other species. The accuracy of calculation by effective binary diffusivity tends to be worse, when the influence of the concentration gradient of the other species is larger than that of its own concentration gradient.

抄録全体を表示

In order to reduce particulate emissions in diesel engines, it is necessary to clarify the fundamen-tal combustion phenomena. In this study, the transient gas diffusion flame was used as a simple and basic model of diesel combustion using acetylene fuel, and the characteristics of the transient gas diffusion flame were elucidated. The measured quantities are the temperature, the relative soot concentration obtained using the two-color method, and the relative concentration of carbon dioxide obtained using the infrared luminous method. The experimental results show that the internal structure of an acetylene gas diffusion flame is composed of four regions ; a combustion promotion region in the upstream periphery, a high temperature and high-soot-concentration region in the upstream inner part, a high-temperature and low-soot-concentration region downstream, and a low-temperature and high-soot-concentration region in the flame tip. Carbon dioxide is mainly generated upstream, and subsequently becomes concentrated in the central part due to the flow.

抄録全体を表示

Characteristics of flame propagation in a vortex core were experimentally investigated Using avortex ring generated by a jet of a mixture ejected through a circular nozzle. The ratio of increasein the flame speed to the maximum tangential velocity Vθmax decreases gradually with increasingVθmax. This is because the radius of curvature of the flame tip in the vortex core decreases withincreasing Vθmax. The flame speed increases with increasing ratio of the densities of the unburned gasto the burned gas(ρu/ρb) 1/2. This is because the radius of curvature of the flame tip increases withincreasing(ρu/ρb)1/2. The flame speed in the vortex core can be predicted well using the model offlame propagation in which the shape of the flame tip In the vortex core is considered.

抄録全体を表示

ombustion characteristics of C3H8 in fluidized bed catalytic combustion, using platinum catalyst mixed with bed material, have been studied experimentally. The fluidized bed material was alumina, in which a supported platinum catalyst was mixed to a percentage weight ranging from 1 to 8. Particle diameter of the bed material was fixed to 0.3-0.355 mm and a static bed height was 20 mm. An increase of combustion temperature or catalyst mixing ratio in the fluidized bed material increases the combustion efficiency. However, the results revealed that catalyst mixing ratio exerts an influence mainly at low mixing ratio. Combustion efficiency almost reaches 100% under the conditions that catalyst mixing ratio was higher than 4% and combustion temperature-was 950K. No NOx formation was detected under these conditions.

抄録全体を表示

In order to reduce noise from an indirect injection (IDI) diesel engine, the combustion, cycle characteristics and a low-noise combustion system were investigated. It was found that realizing a heat release mode in the prechamber which consists of a triangle with a reduced heat release rate are of utmost importance for reducing the combustion-induced noise. Also, it was clarified that a high compression ratio in order to shorten the ignition delay, the control of fuel injection direction in order to restrain air entrainment in the fuel spray, and the enlargement of the jet passage area in order to restrain the swirl velocity in the prechamber are effective as methods for reducing noise. The above combustion system was proven to have a reduced heat release rate, pressure rise in the main chamber and noise level.

抄録全体を表示

The most serious problem affecting diesel combustion is the exhaust emission problem, specifically, of NOx and smoke. Different measures have been undertaken to reduce these emissions, however, a "tradeoff" always exists. EGR is widely known to reduce NO/ emission, but this usually increases smoke. On the other hand, oxygen enrichment reduces smoke, but at the expense of greater NO, emission. This study investigates experimentally the feasibility of the application of high EGR with 02 enrichment in controlling diesel combustion. A condition was found under which the "tradeoff" was nonexistent. The results show the feasibility of some control of exhaust emissions, with respect to smoke and NOx with the application of high EGR and 02 enrichment.

抄録全体を表示