In recent years, development of non-contact fluid control has attracted attention for transporting liquid sample in the fields of biology, chemistry, pharmaceutical and food science. Most of the existing researches deal with the interaction with the non-heated liquid surface and the heated and non-heated solid surface. Although it is of great importance to better understand the droplet dynamics on the heated liquid surface, contactless droplet manipulation on the heated liquid pool has not been well investigated. In this study, we aim to experimentally investigate the levitated and self-propelled droplet on the heated liquid surface with droplet evaporation. In the present paper, we study the effect of the temperature difference between the liquid pool and the droplet and the Marangoni effect associated with self-propelled droplet. This paper describes the critical condition of the levitation and self-propulsion of droplet by visualization of droplet behavior and temperature field.

We measured fluctuations of turbulent shear stress in a horizontal bubbly channel flow. In this research, simultaneous measurement of wall shear stress by a new shear stress sensor using laser Doppler principle and visualization for bubble image by high-speed camera were performed. In order to investigate the relationship between the instantaneous change of shear stress and the passage of bubbles, we evaluated the variation of time series data of the shear stress and the passage distribution of bubbles by analyzing high-order moment and quadrant classification. As a result, it was found that the relationship between the fluctuation of shear stress and the frequency of bubble passage was altered as injection method of bubbles was changed from continuous to repetitive injection. We have confirmed a phenomenon peculiar to repetitive bubble injection, i.e. the shear stress decreased during no bubble passes by. It was also found that the relationship was reversed as the injection frequency of bubbles ranged from 2 to 4 Hz.

Computation of the Basset history force depending on the time history of the relative acceleration between the particle and the fluid at the particle location is very expensive, and hence, in past studies, either the whole part or some fraction of the history has been neglected. Here the effects of the Basset history force on preferential concentration of small particles in a homogeneous isotropic turbulence under a wide range of mass density ratio have been investigated by direct simulations of the Navier-Stokes equations. The Basset force is numerically approximated by the method of van Hinsberg et al. (J. Comput. Phys. 230 (2011) 1465). While the method is found to be fast and accurate, the results obtained from the traditional method taking account of the contributions only over the latest period of a time scale t_win shows a high dependence on t_win. It is clearly revealed that the presence of the Basset force weakens the level of preferential concentration, especially under the conditions of the mass density ratio γ of around 1.5–10 for heavy particles and γ smaller than 0.7 for light particles.

In this paper control of microbubbles in still water is attempted. Microbubbles are generated by electrolysis in water tank. These microbubbles are electrified by cationic surfactant or magnetized by magnetic particles. By applying electric or magnetic field these microbubbles are moved. Influence of applied electric or magnetic field on velocity of microbubble is examined. Electrified microbubbles are successfully moved toward electrode. Velocity of these microbubbles seems almost independent of their diameter. When concentration of surfactant is higher than that of critical micellar concentration (CMC), this trend is remarkably observed. On the other hand it is shown that Magnetized microbubbles don’t move toward intended direction.

In recent years, research on the use of underwater expansion waves in cell engineering and medical applications has attracted attention. Expansion waves can be generated by reflection at interfaces with large acoustic impedance differences, and research on expansion wave focusing using silicon materials is seen. However, there are still many unclear points regarding the characteristics of expansion waves. In this study, we generated underwater expansion wave by underwater shock wave reflection at the water-air interface and the PDMS-air interface. And we focused underwater expansion wave by reflection from an ellipsoidal reflector made of PDMS. The results showed that a large positive pressure appears behind the negative pressure in the expansion wave focused by the PDMS reflector. Then, when the propagation of the internal waves was observed with a two-dimensional reflector, a shock wave along the focusing expansion wave was observed.

Medical mesh nebulizers are used for inhalation therapy of respiratory diseases. In order to effectively manage drug delivery, the size of the generated micro-droplets must be controlled appropriately. As a basic research to develop a medical mesh nebulizer, the purpose of this work is to elucidate the droplet formation process from a 4 micron mesh hole (conical trapezoidal nozzle) of an ultrasonic nebulizer operating at 200 kHz. Numerical simulation of the droplet formation process up to 20 cycles of ultrasonic vibration was carried out for axisymmetric flow using OpenFORM. In the first cycle, no droplets were ejected from the nozzle. In the second to fourth cycles, unstable droplets are formed. After the fourth cycle, two successive droplets from the nozzle merged during two ultrasonic vibration cycles to form a 7-8 micron droplet. The droplets generated after the 4th cycle were found to be stable.

The T-junction is the most common channel shape used for microdroplet generation. To control the droplet size, it is necessary to acquire the time series of each force acting between two fluids. The purpose of this study is to investigate the mechanism of microdroplet formation in micro T-junction using confocal micro PIV (Particle Image Velocimetry). As a result of the measurement, it was confirmed that when the resultant force of the upstream pressure of continuous phase and the shear force acting on the interface exceeded the interfacial tension against the deformation of the interface, it shifted to the Necking stage where the interface spontaneously splits.

The prediction of the flow confined by intricately shaped walls with slip velocity plays a significant role in the design of microfluidic or nanofluidic devices. Immersed boundary methods have been used to simulate flows with arbitrary-shaped boundaries on the Cartesian grid. The key concept of this approach is that the boundary force imposes the no-slip condition which is described with interpolation operator on the immersed boundaries. The boundary force is defined on the immersed boundaries, and smeared on the Cartesian grid by using the regularization operator which composes of the discrete delta functions as well as the interpolation operator. In the immersed boundary projection method which is used for flows over rigid bodies, the boundary force is determined to satisfy the no-slip condition in the same way that the pressure is obtained to satisfy the divergence-free constraint in the fractional step method. We extend this approach to simulations of the flows with partial slip boundaries. In order to impose the partial slip boundary condition on the immersed boundaries, we introduce a new regularization operator which allows the continuous velocity gradient on the immersed boundaries. The channel flow confined by the slip immersed boundaries is computed with the present method.

The resulting velocity profile shows a good agreement with the analytical solution.

Bubble oscillation in the breathing mode at the bubble release was investigated by flow visualization and image processing. The interface amplitude is usually a several micron meter with 1 kHz, so the direct observation is crucially difficult in the experiment. In this report, by means of direct optical observation with the special image processing, the interface displacement has been detected. With a Laplacian pyramid for the captured image group, decomposition, and frequency filtering, a target of fluctuations is able to extract in a sub-pixel order. We verify the method by Wu et al. (2012) and apply it to the breathing mode of bubbles The amplification procedure is as follows: 1) create a hierarchy of image at each time using the Laplacian pyramid method, 2) specify the frequency of the breathing mode, and apply a bandpass filter to each hierarchy, 3) multiply by the magnification and restore the image by overlaying each layer of the Laplacian pyramid, and 4) generate a new image at each time by overlaying the original image. After producing a calibration curve with a numerical simulation, the amplitude of the breathing mode is estimated.

It is known that the surface tension gradient driven by the difference in droplet concentration and temperature causes Marangoni convection. Solutal Marangoni flow has been actively applied in industrial fields such as surface treatment, painting, and inkjet printers. In this study, we focus on the evaporation and atomization behavior of droplets caused by the Marangoni effect for further development of the contactless liquid manipulation. In this paper, the behavior of the binary droplet on the oil pool was visualized and quantitatively estimated the effect of initial concentration, volume of droplet and temperature field.

Reentrant behavior of jet-type cloud cavitation was observed and examined by high-speed observation and image analysis in the flow inside an axis-symmetrical convergent divergent nozzle with different divergent angles. Nozzle A has a relatively small divergent angle, and the nozzle B has a larger divergent angle. For nozzle A, periodic cloud cavitation behavior is observed and reentrant behavior due to pressure waves can be confirmed. On the other hand, for nozzle B there is no periodic behavior as well as no reentrant behavior. It is found that the presence or absence of reentrant behavior of cloud cavitation changes depending on the size of divergent angle.

Oscillatory behavior of the surface of a liquid film on a substrate moving at a constant horizontal velocity is investigated numerically, where the film is confined by a cylindrical solid surface across a gas layer. The evolution of the film surface profile is governed by a thin film equation, in which the disjoining pressure derived from the van der Waals interaction is taken into account. In such a confined geometry, the disjoining pressure acts on the film surface as attractive force from the upper solid surface, and becomes spatially nonuniform because the upper surface is not parallel to the substrate surface. The film surface can oscillate when the substrate moves horizontally at a constant velocity under such circumstances, where the behavior depends on whether the substrate velocity is larger or smaller than a critical value.

We conducted reproduction of the adhering snow shape on a two-dimensional cylinder assuming wet snow using the particle based method and its comparison with the experimental results. As a preliminary calculation, the snow particle trajectory was not affected by the flow in this situation since the Stokes number of the snow particle was around 40. Thereby, only gravity is considered in the following calculation. Next, snow particle accumulation was simulated by estimating both particle number density and inter particle distance. As a result, it was possible to qualitatively obtain the snow accretion shape with surface roughness even when the diameter of the snow particles was small. Therefore, the snow surface becomes rough by applying the particle method. The predicted adhering snow height was in good agreement with experimental results. However, present results could not reproduce the snow glide and creep observed in the actual phenomenon of snow accretion. Therefore, there is a need for further improvements in snow glide and creep. Threshold value of the particle number density for snow accretion judgment affects adhered snow width, while threshold value of inter particle distance affects the snow accretion height and the number of accretion particles. In addition, the effect of changes in particle density with changes in the threshold parameters of particle number density and inter particle distance was also affected.

The fluttering characteristics of a multi-articulated flat plate have been investigated experimentally in a wind tunnel. An articulated flat plate oscillates itself in a mean flow, and it has steady oscillating mode. The flow structures around an articulated plate were three-dimensional. In this report, the pair of tip vorticities were focused. From the results of visualization, it was found that the tip vortices were formed from both of the side edges of the plate and flow down. In this report, to capture these tip vortices, flow visualization and hot-wire anemometry were conducted. In order to visualize the tip vortices, smoke wire method with two wires was used. Hot-wire anemometry with X-type probe was conducted in the cross-section behind the fluttering flat pate. Form the visualization, periodic motion of the moving tip vortices was seen. Furthermore, from the results of the hot-wire anemometry, velocity vectors indicated the rotating flow.

In order to hold two dimensional characteristic in a two dimensional bluff cylinder like a circular cylinder, providing an end plate in the end is known. However, since it strays off from the original purpose about change of the structure of the unification flow after passing over the end plate, it is not interested. So, the knowledge about influence which the magnitude of an end plate does to the magnitude of a bluff cylinder is hardly found. In this study, the flow which passes over a circular cylinder was divided by preparing a partition plate in the circular cylinder span central part. The vortex flow and unification flow visualization observation of each division were performed. The technique of visualization was carried out using a towing water tank by the pouring streak method. The flow velocity is v=0.057 m/s and the Reynolds number corresponding to it is Re=910. It was observed that two flows divided with the partition plate joined at the edge of a partition plate, and formed the complicated flow. Phase difference occurred in formation of the Karman vortex on the back side and the front side of the partition plate. The phase difference is considered depending on the length of the partition plate.

The present paper describes the flow characteristics of air injected from a slit nozzle into a rectangular container. The effect of the slit nozzle offset amount and the jet air velocity on the flow in the container was examined experimentally. The velocity distribution was measured by a particle image velocimetry (PIV) system. When the nozzle is located at the center of the container, the downward impinging jet oscillates largely and frequently in the region of container and does not reach the bottom. When the nozzle is off the center, it goes along the container wall due to the Coandă effect. As a result, the recirculation flow in the container is extended downward. As the Reynolds number (the injected air velocity) increases, two areas of recirculation flow are generated in the upper and lower regions of container. Therefore, the jet does not adhere to the container wall up to the bottom.

This study presents the results of a wind-tunnel experiment on the effects of bar shape on the decay coefficient of the grid turbulence generated by the bars. The present grid bars are set as round bars or square bars following to previous researches. Experimental parameters such as the grid Reynolds number are also set based on previous researchers. Present experimental apparatus consists of a wind-tunnel and a measurement section. A slight diffuser is also used at the downstream end of the measurement section to reduce the exhaust loss. A hot-wire anemometer is used to measure velocity fluctuation of the grid-generated turbulence. The drag coefficient of the turbulence-generating grid is obtained by measuring the static pressure values between upstream and downstream regions of the grid. The mean-fluid acceleration is considered to affect the grid turbulence hardly. A value of the decay coefficient is found to be proportional to the drag coefficient of the turbulence-generating grid.

This study concerns a numerical simulation method of incompressible fluid flows using a weakly compressible explicit scheme and its verification. The Entropically Damped Artificial Compressibility (EDAC) method, one of weakly compressible explicit schemes, is applied for simulations of low-Reynolds number flows with an inlet and outlet. The non-reflective characteristic boundary condition is imposed on the velocity field, and the convective outflow condition is imposed on the pressure field. The Pseudo Crank-Nicolson method with the upstream-shifted interpolation is used for the time discretization. The implemented numerical method is applied to simulate a lid-driven cavity flow and a plane Poiseuille flow. The simulation results agree well with the existing results, then demonstrating the suitability of the method to numerical simulations of incompressible flows.

This study investigates the flow past two cylinders arranged in tandem within a microbubble plume inside a tank. Microbubbles with a mean diameter of 0.055 mm are released by water electrolysis from electrodes placed at the bottom of the tank. Upon rising, these microbubbles induce an upward water flow around them due to buoyancy. Orthogonally to the axis of this microbubble plume, two cylinders with a diameter D of 30 mm are arranged in tandem. The distance between the cylinders, L, ranges between 1.5D and 3D. The bubbles and the water flow around the cylinders are visualized, and the bubble velocity distribution is measured. The experiments reveal the water and bubble shear layers originating at the sides of the lower cylinder, and allow the elucidation of their behavior around the upper cylinder. Furthermore, this study makes clear the effects of L on the flow around the two cylinders, such as the stagnant bubbly flow and the bubbly wake.

To investigate the variation in a sphere's motion and the behavior of the air-water interface induced by the sphere, a solid sphere is launched vertically upward in water toward the air-water interface. Four spheres with different densities ρ_p were studied: a stainless steel sphere (ρ_p = 7930 kg/m³), an alumina sphere (ρ_p = 3900 kg/m³), a Teflon sphere (ρ_p = 2180 kg/m³), and a Duracon sphere (ρ_p = 1360 kg/m³). By experimentally visualizing the sphere motion in water, their launch velocity can be set with high accuracy. When a sphere passes through the air--water interface, it makes the interface more complex and entrains water. Additionally, spheres with a small value of ρ_p were demonstrated to not always vertically collide with the air--water interface.

The purpose of this research is to develop a rated thrust 50kgf class small hybrid rocket engine. In the previous experiment, the injector was damaged due to abnormal combustion in the engine. The cause is that the ignitor was blocked inside the fuel port because the ignitor was larger than the fuel port and nozzle throat diameter. In this study, we investigated the effect of smaller ignitor and a baffle plate installed in the downstream side of the fuel. As a result, the following conclusions. (1) Abnormal combustion didn’t occur. (2) By installing the baffle plate, residential time of paraffin wax becomes longer and promote combustion. (3) The rated thrust reached 50kgf even when the ignitor was smaller. However, in this experiment, the thrust decreased from the middle of combustion time. Combustion is promoted by casting a paraffin wax having a melting point lower than that of acrylic resin. However, since the combustion of paraffin wax ended in the middle of combustion and switched to the combustion of acrylic resin only, the regression rate was reduced and the thrust was thought to have decreased.

Hybrid rocket engine has advantages in cost and safety, compared to conventional type of rockets. The overall objective of this study is to develop and launch a small scale rocket powered by this engine. Also, when launching a rocket, we need to predict the flight trajectory. This is because the rocket should fall into the warning zone decided in advance. A rocket without a control unit for flight trajectory is greatly affected by wind speed and direction against the rocket body. In this study, the flight trajectory of a small rocket was calculated using various drag and lift acting on the rocket, based on the three-dimensional motion model including the influence of the wind. The target altitude of the hybrid rocket under development in this study is 400m. We set launch conditions to achieve this target altitude. The launch conditions are rocket attitude angle, engine combustion time, and natural wind direction. The calculation results indicated a reasonable flight trajectory from the viewpoint of wind effect. In all wind directions, the rocket is expected to land in and near the warning area.

The effect of the outlet shape on a two-dimensional offset jet is investigated experimentally. The offset jet is produced by the flow of air which issue from the end of a long parallel channel. The offset ratio H/h (H : step height, h : channel height) is constant at 2.0, and four types of jet outlet shapes are examined. The exit Reynolds number Re is changed in 2.0 ×10³ to 8.0 ×10³. When the position of the reattachment point is nondimensionalized by the step height, it moves downstream by providing a curvature at the jet outlet in the laminar region. However, the reattachment length is approximately same in the turbulent flow. For all Reynolds numbers, the pressure loss is reduced by rounding corners of the outlet compared to right angles.

This paper focuses on structure of an underexpanded jet spreading radially. The underexpanded radial jet can be seen in various industrial fields, such as intake process of an internal combustion engine, flow rate control by a pressure control valve, molten metal removal in laser cutting process, cooling process of thermal tempering glass and so on. In some of these processes, gas goes through a small gap and radially spreads out at supersonic velocity. In this paper, the underexpanded radial jet is modeled by air discharged from small space between the ends of two circular tubes, which are installed face to face, like a slit-shaped nozzle. Using visualization and acoustic measurement, the underexpanded radial jet from the slit nozzle is examined at different supply pressures. Pictures by shadowgraphy are analyzed to measure cell lengths of the jet, and noises emitted from the jet are analyzed by FFT analyzer to obtain screech tone frequencies. Furthermore, the cellular structure of the radial jet is theoretically analyzed to estimate the cell length. As a result, the cell lengths obtained theoretically agree well with those by the experiment. Furthermore, relation between the jet structure and the screech tone emitted from the jet is discussed based on Strouhal number. It was found that the 3rd cell length plays an important role in the screech noise radiation.

An underexpanded jet impinging on a flat plate is numerically studied in this study. The jet, which is exhausted from a circular nozzle with the nozzle pressure ratio higher than the critical pressure ratio, becomes underexpanded and impinges on the flat plate perpendicularly. Axisymmetric Euler equations are solved using the TVD scheme and the flow field at different nozzle pressure ratios and nozzle-plate distances is simulated to investigate the behavior of shock waves in the jet. At comparatively low pressure ratio, the incident shock wave generated near the nozzle rim intercepts the jet and reaches the jet axis. However, a Mach disk forms across the jet axis at higher pressure ratio and the incident shock does not reach the jet axis but interacts with the Mach disk. A plate shock appears in front of the flat plate, and its location depends on the nozzle-plate distance and the pressure ratio. As a result, it was found that when the plate shock wave is near the node of barrel-shaped cell, the jet oscillates periodically and strongly, and when the Mach disk forms in the jet at high pressure ratio, the periodicity of the oscillation disappears. Furthermore, it was found that the periodicity of the oscillation appears again when the third cell begins to form at larger nozzle-plate distance, even when the Mach disk forms.

In this study, we investigate the thrust force and velocity fields in the wake in a periodic flow for changing pitching angle and frequency of pitching NACA 0012 airfoil. The pitching angle and frequency are changed in the range of 2 to 12 deg and 250 to 750 mHz, respectively. The thrust force and velocity fields in the wake are measured using a load cell and particle image velocimetry (PIV), respectively. For comparison, experiments in a steady flow and a periodic flow are carried out with the same experimental conditions at Reynolds number of 4000. FFT analysis is conducted to investigate the difference in the thrust force spectrum between a steady flow and a periodic flow. It is found that two peaks of the thrust force spectrum occur in periodic flow in addition to peak in a steady flow. As pitching frequency increases, the peak frequency of the spectrum increases. As the pitching angle increases, the peak intensity of the spectrum increases. FFT analysis is conducted to investigate the relationship between the thrust force and velocity field in the wake. It is found that the peak frequency of the thrust force spectrum correlates with the peak frequency of the velocity spectrum in the wake.

There is one of the advanced flow control technique called “Trapped vortex cell (TVC)” control. This technique requires a properly shaped cavity positioned along the spanwise direction on the upper surface of the airfoil. In this study, the final goal is to reveal how the cavity shape is related to the performance of the Cavity wing (referred to as CW). We focused on the two types CW: Closed CW and Opened CW. Between these two CWs, there is a difference in the shape of the trailing edge. We did flow visualization around the Opened CW. We confirmed recirculation flow in the cavity, but it was unsteady. Flow pattern in the cavity was almost same when increasing the angle of attack. We expected that CW has some lift force from 0 to deg based on this flow pattern. We obtained lift curves of these wings by wind tunnel testing. It shows that CWs can get higher lift coefficient than NACA0020 (referred to as NACA) that is the original shape of CWs when the angle of attack is 5 deg. And it was also revealed that there are almost no differences between NACA and CWs lift coefficients when the cavity is the lower surface of the airfoil. But these trends don’t agree with the result of the visualization. We suppose that a difference of Reynolds numbers between the investigation of lift curves and the visualization induced this mismatch. We performed the visualization with a lower Reynolds number because of limits of performance of visualization facilities.

In recent years, EGR coolers, which are heat exchangers for automobiles, are required to improve heat transfer performance and reduce size and weight for the purpose of improving fuel efficiency and reducing nitrogen oxides. Previous studies have revealed that longitudinal vortices induced by vortex generators installed near the surface of the heat exchanger promote heat transfer from the main vortex to the boundary layer. In this study, the heat transfer performance by the shape of the vortex generator (VG) using CFD is compared for the flow in the rectangular pipe channel. Moreover, the effect of the continuous shape is clarified from the calculation result of the case where the VG is installed in the offset fin channel.

The purpose of this study is to obtain drag coefficient of a circular plate with holes. Effects of holes on the drag coefficient are not understood sufficiently yet, however, there are some data about annular plates which show higher value than the circular plate. The drag coefficient shows larger value with larger the ratio of inner and outer diameters. It is strange from the definition of the pressure drag based on the difference between the high pressure in the front stagnation region and the low pressure in the rear separated region. Since there is no difference in pressure at the center of the annular plate because of hole, it is assumed to be a small drag. However, the fact is opposite. The following is a possibility. The test plate is falling in water according to the equation of motion. The change in speed of the plate is measured by a high-speed CCD camera. The experiments were carried out in a water tank. The plate was released from beneath the still water surface, and then, it fell down according to the equation of motion. The drag coefficient Cd of disk with holes is obtained.

The objective of this study is to understand the drag coefficient of concaved solid disc. The concave body is the known as star like shape. The star like shape is the concaved from the pentagonal shape. We are focusing on the shape of concaved from the hexagon. Six types of shape of model was tested. The shape of model considered with the angle of concavity and the circumference of disc. The angle of concave was investigated as 15, 30, 45 degrees. The motion of model was free-fall in a water tank. The tank filled the water. The model motion was taken by using the high speed camera. The image analysis was carried out to investigate the velocity and acceleration of model. The drag coefficient was obtained with the force balance during the motion. The results show that the drag coefficient increase with the increase of the angle of concavity in case of constant projetction area of model. The results with the constant circumference of model show that the drag conefficient increases with the decrease of the angle of concavity of model.

It is well known that the reduction of turbulent friction by adding small amount of the surfactant. Because this method can attain the drag reduction rate up to 80 %, the mechanism of the drag reduction has been studied by many researchers to save the energy. However, there are few studies focused on the dispersion of surfactant solution. In this paper, the effect of surfactant on dispersion of liquid jet is investigated for various concentration. The cationic surfactant used in this study is CTAC and the counter-ion is NaSal at concentration of 30, 100, 200 and 300 ppm at the Reynolds number Re = 700, 1800 and 3500. The vertical concentration distributions of the time-averaged concentration are obtained by adding the uranine to the surfactant solution, showing that the width of jet dispersion is suppressed by surfactant. Moreover, as the surfactant concentration increases, the suppress effect becomes stronger. However, there are no differences at 100 ppm, 200 ppm and 300 ppm. This result indicates limit of the suppression effect. At the low Reynolds number, there is no suppression effect at any concentrations. It indicates that the Reynolds number is important factor to suppress the dispersion of the liquid jet as well as surfactant concentration.

Cavitation is one of the significant issues for water hydraulic components. This study focuses on the relationships between the cavitation phenomenon and noise under various pressure condition. Cavitating jet from a rectangular orifice with different aspect ratio was investigated in this study to reveal the critical parameter for the cavitation in a water hydraulic spool valve. The author performed flow visualization and noise measurement. The experimental results showed that the relationship between jet behavior and noise characteristics. The noises of all aspect ratio jets had a peak at the same positive downstream pressure. The both of cavitation and noise were suppressed for the largest aspect ratio of AR=52.0.

Low-speed streaks over the square arrayed cubes are investigated using large-eddy simulation. A simplified model city was created by arranging blocks in a cube of width H and at a distance of H. Previous studies conducted on simplified arrayed cubes have focused on fully developed turbulent flows. However, the generating mechanism of the low-speed streaks over the square arrayed cubes was unknown. In this study, large-eddy simulation was performed for the transition flow from uniform flow to turbulent flow over the square arrayed cubes to investigate the generating mechanism of the low-speed streaks. The source of the low-speed streaks observed above the cubes was near the ground surface at the second row. This is the position where the influence of the separation flow generated at the first cubes begins to decrease and the inner boundary layer begins to develope. The size of low-momentum fluid increases toward downwind, and is observed as a streak. The occurrence of the low-speed streaks above the cubes was caused by separation flow and ejection of the low-speed streak near the floor. The low-speed streaks fully develop after the 7th row, and the low-speed streak is generated with a certain period.

The effect of regularly arrayed cubes on turbulent flow is investigated using large-eddy simulation. The cubes arranged in a square array are used for an urban canopy to simplify the turbulent phenomenon. In street side, the streamwise flow speed gradually decreases, but becomes constant after the fifth row. In canyon side, a circulating vortex is generated by separation flow from the cubes, and an outflow from the canyon side to the street side is observed near the ground surface. In the turbulent flow after the second row, not only the separation from the cubes but also flow crossing between cubes with a low frequency is generated due to the effect of square arrangement of cubes. The peak frequency of power spectra of the spanwise velocity fluctuations between cubes sifts toward the lower frequency as the fetch increase, and becomes constant after fifth row. This is the same behavior as the decrease in the streamwise velocity at the street side. Turbulent flows linked with street side and canyon side are generated because the peak frequency of the spectra of the streamwise velocity fluctuation at the street side coincides with that at the canyon side at each row. This turbulent flows eventually rise above the cubes.

The mechanism of the low-frequency turbulent flow in an urban canopy formed by series of cuboids is investigated using large-eddy simulation. Three different types of blocks (cubic, vertical rectangle, and horizontal rectangle) was used to investigate the turbulent behavior around series of cuboids in the first, second, third, fifth, seventh and tenth rows. In the vertical rectangle with a double height, the low-frequency turbulent flow is generated at the second row, but in the horizontal rectangle with a double width it is generated at the first row. The peak frequency of the power spectra of the velocity fluctuation moves toward the low frequency as the fetch increases, and the peak frequency becomes constant beyond a certain row. The peak frequency of the energy spectra becomes larger as the block height increases, and becomes smaller as the block width increases. The peak frequency of the power spectra of the spanwise velocity fluctuation between the cuboids coincides with the peak frequency of the power spectra of the streamwise velocity fluctuation side the cuboid. It indicates that the turbulent flow side the blocks strongly affects the low-frequency turbulent flow between cuboids.

Separation of flows from a surface of an object is deeply related to aerodynamics itsself. In this study, a flow around a cylinder is considered as typical flow around the object. When a cylinder rotates, lift force is generated by the Magnus effect, meanwlie at a certain Reynolds number Re and spin ratio, it is reported that a negative Magnus effect (negative lift) is generated by the changes of separation points. And the behavior of the separation point is very important, however there are many unclear points on its fluctuations. Regarding to this, we assume the separations is closely related with the unsteady fluctuation of flow i.e. vortices. So we investigated flows around a cylinder by means of vortex method in order to clarify mechanisms of separations. Computations are performed at Re=2000, and agreements are obtained between the results with vortex method and a summary by Cantwell and Coles (1983). To estimate the location of flow separation we calculated velocity every π/180 rad on the circular arc whose radius is 1.02a (a is radius of the cylinder.). Then the angle θ_s at which the tangential velocity changes from positive to negative was detected on the condition that normal velocity is positive. As a result, the fluctuation of the separation point has been predicted. On the other hand, a relatively large velocity was observed behind the cylinder when vortices leave a cylinder. The influence of the velocity distribution and strength and position of vortices are discussed on the new vortices.

The present paper reports on an experimental investigation of the mixing and development process of multiple elliptic jets with 6×6 square matrix nozzles arrangement at a Reynolds number of 4.0×10³. The multiple elliptic nozzles for aspect ratios of a/b = 2.25 and 6.25 were arranged with the same and the alternate azimuthal orientations. The flow field was measured using the X-type hot-wire probe. The development process of multiple elliptic jets was visualized by the contours of mean and fluctuating velocities at the y-z cross-sections. The multiple elliptic jets merge into a single square jet flow, and the merging point depends on the equivalent diameter. In the case of multiple elliptic jets for a/b = 6.25 with the alternate azimuthal orientation, the axis-switching phenomenon does not occur in the downstream due to the secondary flow induced by the entrainment.

The friction between driving flow and suction flow is known as one of the main loss factors in the ejector. In this paper, the gas ejector efficiency is improved by accelerating suction flow which aims to reduce friction loss between driving flow and suction flow. Design parameters affecting suction mixing section are chosen and 20 cases selected by LHD (Latin Hypercube Design) method are analyzed numerically. 2-dimentional supersonic phenomena including shock wave are simulated. Ejector design is optimized to maximize entrainment ratio with response surface computed by RBF (Radial Basis Function) network. As a result of design optimization, entrainment ratio is 27% improved compared with conventional design. Evaluation of total pressure loss at each section of the ejector concludes that suction flow acceleration reduces loss in the mixing section and improves the entrainment ratio.

Standard ‘static’ square grid is incapable of producing intense turbulence, while active grid produces large deviation in isotropy. In this study, we have developed a method to generate the relatively strong and near-isotropic turbulence 'passively' by tethering a small cylinder in a grid arrangement. Turbulence is found to be enhanced by vortex-induced oscillation due to the small cylinders. Effect of solidity, which is an area ratio of turbulence generator to the test section, on enhancement of turbulence is examined and it is found that larger solidity enhances turbulence more by replacing standard grid for cylinder-tethered grid.

An effective method for the prevention of separation is boundary layer mixing. In this method, the fluid particles in the freestream that have large amounts of energy are supplied to decelerated fluid particles in the boundary layer using longitudinal vortices. Passive control methods such as solid vortex generators are advantageous because of its simplicity, ruggedness, and low cost. However, these methods cannot provide a time-varying control action, and they add parasitic drag in flow situations when the separation control is not needed. On the other hand, there are active control methods that can provide a time-varying control action. A synthetic jet is one of the active control methods. Synthetic jet actuators have advantages in exterior design, safety of vehicles, drag reduction and time-varying control. In this study, the effect of synthetic jets on the drag reduction of the Ahmed body was investigated. Drag measurement was performed by changing the jet-to-freestream velocity ratio and jet positions. In this experiment, two jet positions are tested to investigate the effect of the jet position. The vortical structures around the Ahmed body were visualized by utilizing a smoke wire method. Drag reductions up to 7% were obtained by suppressing flow separation over the slanted surface, and it was found that the control effect was influenced by the jet position.

In order to improve the heat transfer performance of a single impinging jet, multiple impinging jets in which multiple jets are arranged are used. However, the multiple impinging jets has a problem in that the heat transfer characteristics become non-uniform because adjacent jets form a complicated flow field with each other. Therefore, in this paper, in the multiple impinging jet, control is performed to give oscillation to all the nozzles in the row direction at a sin cycle. And it investigated about the influence of the oscillation on the flow structure and heat transfer characteristic of a multiple impinging jets using DNS (direct numerical simulation). In the time-averaged velocity contour, the wall jet spreads in the oscillating direction under all conditions. In addition, strong wall jets were observed when oscillation was applied. The time-averaged Nusselt number contour shows high values near the impinging point under all conditions, and the Nusselt number decreases as going from the impinging point toward the outside of the calculation area. From the RMS value of Nusselt number at each frequency for evaluating the uniformity of the Nusselt number, the uniformity of the Nusselt number was improved by performing oscillation control.

This paper presents the results of numerical simulation of multiple impinging jets. A single impinging jet (SIJ) produces a high heat transfer rate around an impinging position on an impinging wall, while the heat transfer performance decays increasing the distance from the impinging position. Thus, the heat transfer area is narrow and non-uniform heat transfer occurs in SIJ. In order to overcome these problems, multiple impinging jets (MIJ) are generally used. However, MIJ has shortcoming; the interaction between adjacent jets reduces the heat transfer performance on multiple impinging jets. Thus, in order to improve the heat transfer performance in MIJ, it is necessary to develop high-performance-control methods for MIJ. In the previous study, we introduced the intermittent control for MIJ so as to reduce the interaction between each jet. The intermittent control with a phase difference between two jets reproduces intermittently impingement. It turned out that the total heat transfer on the impinging plate is improved under the longer period and large phase difference within the control parameter range. However, it should be considered that the heat transfer performance of impinging jet is influenced by various parameters. In the present study, changing the distance between jets, we evaluate the flow characteristic and heat transfer performance of intermittently controlled MIJ. It turns out that the heat transfer performance of MIJ is improved under the shorter distance between jets and that the wall jet defuses strongly on the shorter distance.

Since a fan cooling for electronic devices is conventionally used, it is important to understand the flow phenomena in order to achieve efficient cooling performance. In this study, we investigate the heat-flow characteristics of around a single and two small-axial fans under the influence of varied the distance between centers of two fans over the parallel-plate having a heating wall using LES (large eddy simulation). An IB (Immersed Boundary) method is introduced to reproduce the impingement wall, small-axial fans and casings. We analyze the unsteady flow and heat transfer in a wake of the fans.

A wide variety of applications of jets enhancing mixing, diffusing, heating and cooling efficiency have been studied. For example, jets are used in the industrial processes such as combustion, ventilation and so on, and their improvement of mixing and diffusion is demanded. It is well known that the characteristics such as mixing and diffusion of the jet are markedly dependent on the large-scale vortex structures being formed near the nozzles. Therefore, as giving appropriate inflow conditions at a nozzle, it is capable of controlling the large vortex structures near field around nozzle and is expected to improve the mixing performance. In particular, blooming jets occur with superimposition of axial and helical excitations on the inlet velocity profile. Blooming jets are characterized by vortex rings moving along branches of separate streams. In previous studies, it is observed that blooming jets change the number of branches with specific frequency ratio of axial to helical. The present study evaluates the mixing performance of blooming jet using direct numerical simulation (DNS) varying the frequency ratio. It reveals that the statistical entropy and entrainment rate are obviously greater than that of unexcited jet as well as the momentum transfer of blooming jet is significantly enhanced.

To investigate the change of translational velocity of negatively buoyant vortex rings, PIV measurement has been performed with four high-resolution cameras placed in a line in the direction of the ring movement in the range of Reynolds number based on a jet velocity, 10,000 < Re_Ujmax < 25,000 and Froude number, 3.0 < Fr < 16.8. It is found that the vorticity distribution in the core is nearly Gaussian to be the same as non-buoyant vortex rings and the decreasing rates of circulation and radius of rings with time are proportional to 1/Fr². The impact of the decrease of the translational velocity from the variation of circulation, radius of ring, and radius of core is discussed by using theoretical formula for a thin vortex ring.