We investigated the influence of the wall surface properties on the drag reduction and turbulent transition from the viewpoint of the surface flexibility. The flow rate and pressure loss were measured for the water flows in a rectangular channel with an aspect ratio of ten. The test surfaces were the smooth surface, home-made fiber-flocking (flexible riblet) surface, and the 3D-printed rigid riblet surface. The flexible and rigid riblet surfaces have the same shape. A confocal laser microscope confirmed that the flexible and rigid riblet surfaces have almost an identical dimension. As a result of the present experiment, we mainly obtained following findings. First, the flexible riblet surface caused the delay of turbulent transition, while the rigid one didn’t. Second, it was found that both surfaces showed almost the same maximum drag reduction ratio, but the flexible riblet surface had the smaller drag reduction region of the Reynolds number than that of the rigid one. Third, the roughness effect of the rigid riblet surface was larger than that of the flexible one at the higher Reynolds number.

Flows in a corrugated channel with periodic heating were examined experimentally. The channel was set horizontally, with its both ends being opened to the laboratory. The upper wall was smooth and the lower one corrugated whose amplitudes and wavelength was 0.1h and 6.28h, respectively, where h denotes the channel half-width. The upper wall was kept isothermal at the middle temperature of periodic heating, while the lower one was subjected to a sinusoidal temperature distribution by repetition of heating and cooling. The Rayleigh number based on the peak-to-peak amplitude of periodic heating with the same wavelength as that of the corrugation was Ra ≦ 3000. The phase difference between heating and corrugation was ±π/2. The results showed that pairs of counter rotating rolls accompanied with a meandering stream tube were realized, generating the mean horizontal flow. The direction of net flow was reversed when the heating phase was shifted by 180 degrees. It was also found that net flowrate increases monotonically with the Rayleigh number examined.

From the effective use of energy, it is desired to develop a vortex generator that realizes heat transfer enhancement while suppressing flow resistance, that is, the dissimilar heat transfer promoter. In recent years, various thermal-fluid engineering devices such as heat sinks have been developed in laminar region by topology optimization using adjoint analysis. However, in a strong non-linear flow field, second or higher-order variation cannot be ignored if integration is performed for a long time, in other words, the variational principle breaks down. Hence this optimization method cannot be applied to the development of dissimilar heat transfer promoters. Therefore, in this study, we develop a dissimilar heat transfer promoter by genetic topology optimization called the NG-net method, which is the method without using gradient information of objective functional. The purpose of this study is to find out what shape vortex generators should be installed in order to maximize heat transfer in plane channel with temperature differences on the upper and lower walls. The Reynolds number is set in a range where vortices appear time-dependently. In the poster presentation, the optimum shape, the Reynolds-number and time-integration-dependence of this shape will be shown, and the reason why the obtained shape realizes greater heat transfer will be discussed.

Skin-friction drag reduction by combination of riblet surface and large-scale control of fully developed turbulent channel flows is investigated by direct numerical simulations. To discuss the effect, flow statistics and instantaneous vortical structures are shown. It was found that the drag reduction effect is enhanced by the combination of control techniques: vortical structures are generated by the Large-Scale control locally, but these are pushed-up by the riblet surface.

Convective heat transfer of a pipe flow with vibration control is investigated by direct numerical simulation (DNS). This study aims at the application like a catalytic converter, which consists of thin tubes. The effect of spatial development in the flow and thermal field toward the downstream was examined. The computational domain consists of two parts: a main part for vibration control and a driver part for generating turbulent inflow of the main part. The friction Reynolds number Re_τ is 110 in the driver part and 60 in the main part. Vibration control is expressed as inertial force applied to the fluid, and its parameters are the amplitude and the frequency. Owing to the vibration control, the Nusselt number Nu increases almost all over the domain than that without control. In the downstream part of the domain, Nu with vibration becomes almost constant and is similar to the value in a previous study applying periodic boundary condition in the axial direction. Increasing rate of Nu in the downstream part is 53%. The radial profile of radial turbulent intensity and radial turbulent heat flux becomes converged with spatial development. The phase-averaged temperature field moves perpendicular to the axis according to the vibration phase and approaches the wall. The movement in the downstream part is not changed in the axial direction.

Opposition controlled fully developed turbulent flows along a thin cylinder are performed by means of direct numerical simulations. An influence of cylinder curvature on the skin-friction drag reduction effect by the classical opposition control (i.e., the radial velocity control) is mainly investigated. The curvature of cylinder affects the statistics of the flow; however, the control effect in the small curvature case is similar to that in channel flow. For large curvature, the drag is reduced even though the detection plane is located away the wall surface. The reduction rates of the present cases are larger than those in the channel flow.

Interactions between fluid flow, in particular, turbulent flow, and solid particles are complex because many parameters control the interactions. In experiments of particle-laden turbulent flow, it is difficult to identify the scale of vortices which interacts strongly with the particles, since vortices of various scales coexist in turbulence. We, therefore, focus on the effect of the addition of solid particles to a single-scale vertical flow. More concretely, in our experiments, we use a container with a spherical cavity of 180 mm diameter. The container is filled with water and many rigid rod-like particles (the length is 12 mm and the diameter is 2 mm) are laden. We rotate the container at a constant angular velocity to create a swirling flow. This flow would be the solid-body rotation if the particles were not added. Our experiments show that the flow velocity in a central region of the cavity is drastically reduced when we add particles with a density slightly higher than that of the working fluid. We aim to explore the physical mechanism of this suppression phenomenon. For this purpose, we quantify the suppression rate by using particle tracking velocimetry (PTV) and reveal the condition when the addition of rod-like particles suppresses the swirling flow.

Drag reduction effect of traveling wave control in turbulent pipe flow is experimentally investigated. A pipe is made of a rubber sheet in order to obtain a large wall deformation amplitude and it is set vertically. Three piezoelectric actuators are mounted at the upstream region of the test section as the vibration sources to produce downstream waves. Wave parameters show effective values for drag reduction by adjusting input parameters of the actuators. The drag reduction rate of 5% is obtained at the bulk Reynolds number of 4800. Laser Doppler velocimetry measurement is performed in order to evaluate the control effect on the flow field. The profiles of mean streamwise velocity and root mean square (rms) value of the streamwise velocity fluctuations in the uncontrolled case are in good agreement with those of DNS. When the traveling wave control is applied, the mean streamwise velocity gradient becomes smaller near the wall, which indicates the drag reduction effect. Rms values of the streamwise velocity in the controlled case is larger than those of the uncontrolled case. By conducting three-component decomposition, it is observed that a periodic component is generated near the wall, whereas a random component decreases compared with the uncontrolled case. Thus, the influence of the traveling wave control on the turbulent flow field near the wall in a pipe is experimentally confirmed.

We carried out direct numerical simulation (DNS) of pulsating turbulent pipe flow and predicted the time evolution of the flow field by deep learning. Pulsation control is one of the turbulence control methods to realize drag reduction. A rapid prediction of pulsating flows by deep learning is beneficial to optimize the pulsation patterns. The deep learning model in the present study consists of a convolutional autoencoder (CAE) and two-layers of long short-term memory (LSTM). The training data are images of velocities and pressure field in a cross-section which is calculated by the DNS. The feature vectors are extracted from the images by the CAE. The time-dependency of the flow dynamics is learned by the LSTM. The spatially averaged pressure gradient that is given to pulsate the flow is concatenated to the feature vector. By applying sequence-to-sequence learning to the model to predict the long-term dynamics, the model successfully reproduced the time evolution of the distribution of velocity and pressure fields and the flow statistics. The temporal variation of the friction coefficient calculated from the predicted flow field is approximately identical to those of the DNS. The relative error of the time-averaged friction coefficient is 6.0%. The model roughly predicted the friction drag of the pulsating flow.

Injection of bubbles into a turbulent boundary layer has multiple impacts. And frictional drag reduction is a phenomenon from their combined effect. In numerical way for drag reduction by injecting bubbles, recent works have been targeted small bubble which is below Weber number 1. Although several systems have been developed based on these bubbles, it has some difficulties in applying in practice. Thus, to solve this problem, we approached from other aspects, such as relatively large bubbles. Thus, the research objectives are checking the capability of simulation on large bubble, which is high We number over 100, and understanding its characteristic. In the case of numerical condition, it was divided into 3 stage. (generating turbulent channel flow, Injecting bubble, simulation on turbulent channel flow with large bubble) The numerical results exhibit good agreement with experimental results on the bubble shape.

We performed direct numerical simulations of turbulent boundary layer flows of viscoelastic fluids with taking into account the additive concentration variation where the viscoelastic behavior is represented by using the FENE-P model. In the present simulations, the initial and boundary conditions were set such that the additives were injected from the wall surface. We compared the diffusion processes between Newtonian and drag-reducing viscoelastic fluids for the Schmidt number Sc = 1, 5 and 10 at the momentum Reynolds number Re_θ0 = 670. In addition, the difference in local drag reduction ratio along the streamwise direction between homogeneous and inhomogeneous viscoelastic fluids was investigated at the Weissenberg number Wi = 50.

It is well known that a small amount of polymer additive mixed into Newtonian fluid can reduce the frictional drag of wall turbulence. Such a fluid with the drag reduction effect is a viscoelastic fluid, and the prediction of viscoelastic turbulent flow by RANS (Reynolds-Averaged Navier-Stokes) simulation is required for engineering applications. In RANS to predict the viscoelastic turbulent flow, three unknown terms are involved into the ensembled-averaged constitutive equation. In the flat channel flow, Λ_ij is dominant compared to the other terms and they except for Λ_ij are neglected. However, the comparison of the three unknown terms with different analytical configurations was not performed. Therefore, we performed DNS of viscoelastic turbulent flows over a backward-facing step and investigated the influence of the three unknown terms. In the result, the turbulent diffusion term Ξ_ij is comparable to Λ_ij near the separated shear layer due to the separation. So, a model of Ξ_ij should be constructed considering the result that has a noticeable value in separated shear layer for improving the prediction accuracy of RANS.

We performed stereoscopic PIV measurements for the drag-reducing turbulent boundary layer flows by injecting surfactant aqueous solution. For the case of the drag reduction ratio more than 60%, large-scale turbulence structures were often observed in the fluctuating velocity field. In this study, we discussed the identification of such large-scale turbulence structures by using the second invariant of the velocity gradient tensor and the imaginary part of the eigenvalues.

The present study focuses on the acoustic streaming generated by an ultrasound transducer in the air. A commercially available ultrasound transducer was operated by the sine wave supplied from the function generator at the resonance frequency of 40 kHz, which was amplified by a high-speed voltage amplifier. The distribution of sound pressure fluctuation was measured using a microphone. The velocity distribution was evaluated using PIV. The distribution of root mean square value of sound pressure fluctuation mostly indicates an axisymmetric distribution. The sound pressure rms value increases with the input voltage amplitude. The flow velocity increases with the input voltage amplitude since the driving force is proportional to the variance of the pressure fluctuation. The relationship between the induced flow velocity and the sound pressure fluctuation level has been clearly presented. When the driving force is large, the flow apparently exhibits unsteady behavior.

The heat transfer enhancement is of importance in the low Reynolds number flow, because small heat exchangers with narrow flow passages are required recently. In this study, we investigate the heat transfer enhancement effect by ordinal feedback control techniques in a channel flow by means of direct numerical simulations. As control technique, the blowing and suction are imposed based on the near-wall velocity and the control is designed to promote turbulence and enhance the heat transfer. As results, the skin-friction drag and heat transfer are sustained at the fully turbulent level when an amplitude of the blowing and suction is small. On the other hand, these are larger than that when the amplitude is comparable to the near-wall normal velocity.

In this study, multiple-passage channel flows are investigated experimentally. The multiple-passage is a parallel flow type, and consists of the five branch channels. The dielectric barrier discharge plasma actuator (PA) is used for the flow control. The location in which the actuator is installed is changed in three positions of the outside wall at the inlet manifold and the outlet manifold. Reynolds number based on the bulk velocity and hydraulic diameter at the inlet manifold channel is varied from 6.0 × 10² to 2.0 × 10³. The results show that the pressure loss decreases and the flow rate in each branch channel passages improves by the effect of induced flow with the PA. The effect is remarkable when the PA is installed near the first branch of the outlet manifold. There is an appropriate PA drive frequency depending on the Reynolds number.

The wing-tip vortex is closely related to the induced drag of airplane, and its suppression is desired to improve flight efficiency. In the present study, the decay of the wing-tip vortex is enhanced by introducing wing-tip vibration. A vibrating wing-tip is attached to a NACA0012 half wing model. The stereo PIV measurements at z/c = 4 have been performed to evaluate the velocity distribution of the vortex. The vibration frequency, f, and the angle of attack of the wing, α are taken as parameters for the present experiment. The effect of the wing-tip vibration is observed both in the axial and in-plane velocity distributions. For the stationary case at larger angle of attack, α = 7 deg and 10 deg, the tangential velocity of the vortex exhibits a distinct peak in its radial profile. This peak disappears with the wing vibration indicating that the core of the wing tip vortex is weakened by the vibration. It is confirmed that the vibration of the wing-tip exerts significant impact on the development of the wing-tip vortex.

In order to clarify effects of electrode shape of the plasma actuators (PA) on the flow control for an oscillating square cylinder, wind tunnel experiments were conducted. The control effects of the separation flow were evaluated by flow visualization with a smoke-wire method and wake flow measurement with a hot-wire anemometer. As a result, the suppression effect of the separation boundary layer around the stationary cylinder depends on the electrode shape of the PA. In this experimental condition, the opposite arrangement of introducing vertical vortices into the separated boundary layer has the biggest suppression effect, which reduces the thickness of the separated boundary layer to 37%.

It was also confirmed that flow fields around the oscillating cylinder can be controlled by using PA. The vortex shedding frequency associated with the oscillation can be halved and the velocity fluctuation intensity can be suppressed by 24%.

Since penguins prey underwater animals such as krills and small fishes, it must be important to minimize cost of swimming for the penguins. Their body surface was covered by aligned feathers of which barbs (i.e. branches of a feather) forms micro grooves, that perhaps influences friction drag of the body. Moreover, the feathers possibly affect the boundary layer separation from the body. Here we investigated drag reduction effect of micro grooves mimicking the penguin’s feathers by water tunnel experiments. Firstly, we made a model of the penguin body that reproduces the dorsal curvature of a Gentoo penguin from a side-view image. Secondly, we created micro-grooved film using a UV-laser cutting machine. The size of the micro groove was determined based on the measured morphology of real feathers and estimated thickness of viscous sublayer assuming a flat-plate theory. Then the micro-grooved sheet was affixed to the body model, and its drag was conducted in a water tunnel. As a result, it was found that the maximum drag-reduction ratio was about 3.8% when the spacing of the groove, s, was 0.1 mm and flow speed was 2 ms^-1. The model with s of 0.055 mm, which corresponds to the real feathers, achieved drag-reduction rate of 1.9% at flow speed of 2 m s^-1. Therefore, it can be said that the penguin’s feathers have reduced drag. This drag change was bigger than the effect of riblets on friction drag. This may be due to imitation surfaces affecting boundary layer separation.

In this study, a consideration for measurement and control of very slow airflow under painting environment was carried out to generate the flow field without dust from the booth. The control of flow direction was studied by using multiple counter-rotating fans that can generate fine uniform flow over long distance. The counter-rotating fans were controlled for the rotational speed, and the generating air flow was measured by using MEMS flow sensor and thermal anemometer. The simple method by using MEMS sensor, that has high directivity, was discussed to detect the flow direction. It was found that the MEMS sensor has high time response in comparison with the thermal anemometer, although the velocities obtained from two sensors were almost the same under the steady state. The flow direction can be changed by using the counter-rotating fan without the wind tunnel for reduction of turbulence. Experimental values related to the flow direction were obtained by using multiple MEMS flow sensors. The results suggest that MEMS sensors should be used in vertical mounting. It is possible to evaluate the relationship between the angle and the velocity, can be useful to develop the simple method for the detecting flow direction in near future.

Recently, the effect of drag reduction has attracted a growing interest from the stand point of energy conservation. In this study, the effect of a water-shedding coating material consisting primary of SiO2, applied to the surface of a flapping wing is investigated by means of aerodynamic force measurement and PIV analysis.

Balls used in ball games have various surface structures that affect aerodynamic characteristics. In soccer, a new ball is developed for each global competition, and the aerodynamic characteristics differ because the panel shape and surface structure of the balls are different. In the author's research, the aerodynamic characteristics in the non-rotating balls were evaluated using a model ball in which the groove depth and groove width were systematically changed. However, the ball in the actual game is rotating, but the aerodynamic characteristics in the rotating state to the aerodynamic characteristics are still unclear. In this study, the effects of groove depth and groove width on a soccer ball in a rotating state are clarified.

In recent years, soccer balls of various designs have been developed, but depending on the design of the "panel" on the ball surface, there is a case that a "Knuckle ball" in which the trajectory changes irregularly during flight may occur. Knuckle ball is an element that audiences enjoy in games, but it is an obstacle in fostering players. In this research, we aim to clarify the influence of the panel on the surface of the soccer ball on " Knuckle ball" by wind tunnel experiment and to grasp the relationship between the panel shape of the ball and aerodynamic characteristics.

Soft tennis is played using rubber balls. During the competition, it basically flies with top spin, so it draws a parabola with gravity and Magnus force. However, balls that takes up a rising trajectory, peculiar flight, appears unexpectedly. It is because of the negative Magnus effect. Therefore, the purpose of this research is to observe the range of this negative Magnus effect and to investigate the force occurring in a peculiar flight aerodynamically and the state of the flow. PIV measurement was conducted to analyze the flow around the ball when negative magnus effect occurred. As a result, it was found that the Magnus force and the negative Magnus force are related to the flow behind the ball. In addition, it was confirmed that the speed of the vertical speed of the ball was switched. As a result, the position of the separation point changed and it was found that a negative Magnus force was generated.

In this study, we experimentally and numerically examine a three-fluid flow that a single falling drop collide with immiscible liquid layers. In our study, silicon oil is used for drop and three kinds of syrup solutions with different viscosity values are used for immiscible liquid layer. In our computations, a Moment-of-Fluid (MOF) method is used to capture the interface of three immiscible fluids. From experimental and numerical results, the formation of jet is observed when a drop impact on the surface of a liquid layer with lower viscosities. Meanwhile, the jet is not formed when a drop impact on the surface of a liquid layer with higher viscosity. Our numerical results agree well with experimental ones: numerical results successfully predict the complex fluid dynamics value that a drop collides with an immiscible liquid with free surface. The MOF method allows us to capture the dynamics of three-fluid flow.

Researches on free fall of liquid droplets have been conducted for a long time. And in these days numerical simulations of two-phase flow are carried out. However, there are few numerical analyses of long term free fall of a liquid droplet. In this study, we tried to reproduce the free fall phenomenon of a liquid droplet by numerical analysis. A two-phase flow analysis program was developed based on CCUP method and Level Set method. And the numerical analysis of a droplet falling in still air was performed. The droplet diameter was 2 mm. Analysis is performed on a supercomputer, using MPI parallel and OMP parallel. The falling velocity of droplet converged to 3 m/s as terminal velocity. The droplet showed a vertically stretched shape at terminal velocity. These are inconsistent with observations by previous researchers, the terminal velocity is about 6 m/s, and the droplet shows an umbrella shape.

Clustering of spherical bubbles rising along an inclined flat wall has been investigated experimentally to analyze the bubble-bubble interaction. To avoid bubble coalescence, electrolyte (MgSO₄) solution and surfactant (Triton X-100) solution are utilized as a liquid phase. In MgSO₄ solution, free-slip boundary condition of bubble surface is maintained, on the other hand, the boundary condition of the surface of rising bubbles in Triton X-100 solution becomes no-slip due to well-know Marangoni effect. The development of the bubble clusters and the bubble-bubble interaction of two, three or four bubbles are discussed through the quantitative evaluation of the bubble velocities and the distance between neighboring bubbles. The bubbles tend to be arranged horizontally as they rise, which induces bubble cluster formation. When two bubbles are arranged vertically, a trailing bubble is accelerated by the wake of a leading bubble. After this stage, these two bubbles take side-by-side arrangement. In the development of bubble clusters, the horizontally aligned bubbles tend to rise more stably in case of the no-slip condition. The drag coefficient increases with increasing number of horizontally aligned bubbles. The decrease in the rising velocity becomes less in the case of no-slip condition. The distance between neighboring bubbles also decreases with increase in the number of horizontally aligned bubbles, and this distance tends to be larger in case of no-slip condition.

The pressure measurement using a fiber optic prove hydrophone (FOPH) and the observation with a high-speed video camera have been conducted simultaneously to investigate the collapse of a laser-induced bubble near a rigid wall and the corresponding shockwave generation due to the bubble collapse. Prior to the experiment of the bubble collapse near a rigid wall, the experiment of a laser-induced bubble in infinite liquid has been conducted to check the applicability of the FOPH to the pressure measurement of shockwave generated by the bubble collapse. The directivity of FOPH is evaluated by controlling a directional angle to the spherical shockwave and a distance from a collapsing bubble. Although the peak value of the pressure decreased with an increase in the directional angle, only 5 % reduction of the peak value is confirmed in case of 20 degrees of the directional angle. The peak value of the shockwave decays in proportion to the -1th power of the distance, which shows the characteristic of a spherical wave propagation. In the experiment of the collapse of a laser-induced bubble near a rigid wall, the shockwave generated from a collapsing bubble have double peaks due to the micro-jet penetration and the bubble rebound. The first pressure peak corresponds to the micro-jet penetration and is proportional to the stand-off parameter from a wall and also correlates well with the water hummer pressure evaluated form the tip velocity of a micro jet measured by the high-speed observation.

The drainage and rupture of the thin liquid film formed between a flat mica plate and an approaching bubble with a constant velocity have been investigated experimentally. The effect of impurities in water on the drainage and rupture of the liquid film has also been evaluated by the comparison of the solutions: ultrapure water, two electrolyte solutions (0.05 M MgSO₄ solution and 0.05 M CH₃COONa solution) and surfactant solution (10 ppm Triton X-100 solution). The approaching velocity of a bubble toward a mica plate was controlled in the range from 1 μm/s to 1000 μm/s and the development of the relative thickness distribution of a thin liquid film was measured with the interferometer and the high-speed observation. As a bubble approaches toward the mica surface, the film shape changes from pimple to dimple, and the depth of dimple does not change during the dimple formation and the rim of the dimple spread wider. By the increase of approaching velocity, the depth of dimple tends to be deeper. By the comparison of 0.05 M MgSO₄ solution, 0.05 M CH₃COONa solution and ultrapure water, the dimple formations until the thinnest film thickness at the rim of the dimple becomes less than the limit of the thickness measurement of about 100 nm are similar. On the other hand, the timing of rupture retards in the electrolyte solutions, and its tendency is qualitatively consistent with the intensity of the effect of the bubble coalescence inhibition. The film with a thickness of less than about 100 nm becomes more stable when we used the mica plate as a solid surface and the rupture timing remarkably retards comparing with the case of a glass plate.

The gas-liquid two-phase flow in a simplified evaporator that has multi-pass channels and cool the air as an automobile component was numerically analyzed. One of the issues was evenly distributing refrigerant to channels. In this study, we tried simulating air/water two-phase flow instead of refrigerant two-phase flow by particle method that represents air and water with particles. To solve this incompressible air/water two-phase flow, this study used the moving particle semi-implicit (MPS) method that is one of particle methods. Numerical reproduction was intended to form small water droplets in channels, splash, clear air/water interface and water flow distribution into channels. We found that collision distance and density difference between air and water particles remarkably influenced on forming air/water two-phase interface. Small water droplets, splash and air/water interface were formed clearly. In this simulation, large collision distance enough to move gently and pressure relaxation caused by density difference both contributed to the above objective result. In addition, water tended to flow in the upstream channels and reduced gradually in the direction of the downstream as the same result as the experimental results. Downstream waves were numerically produced by collision between upstream and downstream water every 0.4 second. The period was almost the same as the experimental period of 0.3 second.

Level set methods and volume of fluid (VOF) methods are popular for computation of immiscible-fluid flows. Here the ACLS (accurate conservative level set) method, a level set variant, and the CICSAM (compressive interface capturing scheme for arbitrary meshes) method, a VOF variant, are applied to the Zalesak’s problem (solid body rotation of a slotted disk) to clarify the pros and cons of these methods. The ACLS combined with a high-order scheme shows a remarkable ability for preserving both the shape and volume of the disk. While the CICSAM presents a severe restriction on the time increment, the impact of the difficulty can be relieved by introducing a so-called sub-stepping strategy in practical applications and hence it is also a useful methodology due to the simplicity.

We attempted perfect gas discharging from closed end holes in liquid due to acoustic wave irradiation. An underwater speaker was set in front of test samples in liquid pool and pressure waves of sinusoidal wave, which have constant frequency or whose frequency changes in time, i.e. sweep wave, were irradiated. We observed the gas discharging process with a high-speed video camera during acoustic wave irradiation to investigate the factors that promoted gas discharging. As a result, we found that the sweep wave irradiation was more effective for gas discharging than that of constant frequency and the sweep wave was able to achieve complete gas discharging with several seconds. In addition, we found that the discharging process consists of three stages. The detail of the three steps are discussed.