The mechanisms of chromosome condensation and decondensation processes plays a key role in regulating transcriptions and gene expressions, but it is not sufficiently understood, especially in human cells. So far, three types of chromosome models have been proposed and widely known, while recent experiments have shown that there is polymer-melt-like state which is absent in those models(1). These indicates that the fluidic and density effects by crowded polymers should be taken into dynamical models explicitly in order to explain the above dynamical nature. One of the prototypes for such models is the Rouse model with the Brownian dynamics method. But the method comes with a short time step (1ps ~ 1ns) for its constant force assumption. In this study, we aim to realise in-silico chromatin dynamics in human cells by improving the Brownian dynamics for longer time intervals than picoseconds, and by implementing the density effect as collisions. To the generalized collision process, we simulated the motion of two nucleosomes including the LJ potential and extracted their existence probabilities as discrete data from conputer experiments. Furthermore, by increasing the number of samples, we examined what arbitrary probability density function we could converge to. As a result, the particle movements imply a certain distribution with two peakes whose gap is about the particle radius in the perpendicular direction to the initial to the initial axis of the pair.This suggests the collision process may be generalized.

This study focuses on the relationship between the fiber orientation behavior and the solvent viscosity for fiber orientation and the parameters of the F-T-S model of Rheological model. The step shear experiments applied to fiber suspension with different solvent viscosity. The fiber orientation was visualized by a microscope attached to a rotational rheometer with glass parallel plate flow cell. As an experimental result, the viscosity of solvent is independent to the steady state of fiber orientation, but not to the transient state. Then, the visualized orientation behavior was fitted by the orientation predicted by the F-T-S model. The two model parameters, the F-T constant and the slip factor, respectively have the relationship to the steady state and the transient state of the fiber orientation. As the same as the experimental behavior, the slip factor dramatically depends on the solvent viscosity. The F-T constant is agreed with the equation by Phan-Thien et al.

The viscoelastic fluid flow in a rotating double cylinder was investigated by numerical simulation using a finite-difference method. The volume penalization method was used to represent both a rotating concave-convex inner cylinder and a static outer cylinder. The governing equations consist of the stream function-vorticity formulation of the momentum equation and the constitutive equations of the UCM (Upper Convected Maxwell) fluid. The results show that, in steady states, the circumferential velocity is higher than that predicted by the Couette flow solution, with the effective radius of the inner cylinder increasing accordingly. In the vorticity distribution, ring-like high vorticity areas corresponding to the sudden changes in velocity are observed under higher rotating speeds.

Viscoelasticity of a dilute surfactant solution can drastically change the turbulent flow of water. This results in many differences in turbulent transport phenomena. We experimentally investigate the differences in the transfer of angular momentum in turbulent Taylor-Couette flow with an ultimately high Taylor number regime. We report the angular momentum transfer is suppressed in the counter-rotating regime. Furthermore, we found remarkable the enhancement of the wall frictional drag in the co-rotating regime while drag reduction was observed in the counter-rotating regime. Interestingly, a non-monotonical profile of the mean azimuthal velocity appears for the drag enhancement cases. The drag enhancement and non-monotonical velocity profile are coincident with a previous numerical investigation whereas some control parameters are quite different from our experiments.

We applied a deep learning to construct a surrogate model of the constitutive equation to simulate the viscoelastic turbulent channel flow. We used U-Net, a convolutional neural network (CNN), to predict the conformation stress field from the instantaneous velocity field obtained from direct numerical simulation (DNS). We investigated the prediction accuracy of the conformation stress and the possibility of the surrogate model of U-Net as a DNS-CNN simulation. The mean profile of c_xx and the flow field were reproduced with high accuracy by using U-Net. However, the prediction accuracy of c_yy in the buffer layer, where polymer rotation is induced by vortices, is low. The predicted instantaneous fields exhibit noise caused by the convolution operation of U-Net. In the surrogate modeling of U-Net, we compared the result of the DNS-CNN simulation to the pure DNS. After 5000 steps, the instantaneous flow fields were slightly different, but the obtained statistics showed a good agreement between both simulations.

The explosive vortex stretching of the thread parts in vortex reconnection phenomenon is considered to be related to the occurrence of fast energy cascade in turbulence. In this study, we investigated whether the vortex stretching in the reconnection can be controlled by artificially multiplying the vortex-induced velocity by a control coefficient in numerical simulation. As a result, it was found that the vortex stretching could be suppressed successfully by suppressing the induced velocity. This result provided a technical suggestion for controlling the occurrence of energy cascade in turbulence.

In general, the vortex motion is described by the Biot-Savart law. Furthermore, under the localized induction approximation (LIA), the curvature and induced velocity of vortex are proportional, and then the vortex motion is described by the nonlinear Schrödinger equation. In this study, we focused on the vortex reconnection, which is one of the vortex interactions. In the reconnection problem of two vortex rings, by numerically analyzing the relationship between the curvature and induced velocity along the vortex ring, it was investigated at which location of the vortex ring the LIA held and at which location the LIA did not hold. The results helped a theoretical understanding of the reconnection.

Turbulent incompressible flow through a pipe with a rectifier part composed of 19 narrow tubes is calculated by Large Eddy Simulation (LES) with the immersed Boundary (IB) method and the effect of the rectifier-part length on the turbulent field is studied in the present work. The streamwise mean flow accelerates near the center of the pipe in the rectifier tubes. As a result, turbulence is suppressed by the tube bundle rectifier in its downstream region and the suppression is strengthen as the length of the rectifier part increases.

In order to investigate the effect of geometrical characteristics of rough surfaces on turbulent frictional resistance, we performed particle image velocimetry (PIV) measurements on the turbulent flows over systematically varied rough surfaces. The present study focused on two characteristics parameters, namely the skewness factor and effective slope, which are respectively related to the probability density of roughness elevations and the wavelength of the surface undulations. It is found that the frictional resistance is large for the surface with peak-dominated steep rough surface, which is characterized by the positive skewness factor and large effective slope value. The characteristic parameters are found to have a minor influence on the asymptotic behavior toward the fully rough regime.

The effect of roughness pitch ratio for a wall-bounded shear flow has been investigated based on spatial averaging operation of the Reynolds equation for a channel flow. The roughness pitch ratio was PR =1, 2, 4, 8 and 16.The velocity field was measured with 2-D PIV system. The Reynolds number based on the channel height was 2500. The spatially averaged streamwise mean velocity profiles of PR = 4, 8 and 16 were defected compared to those of PR =1 and 2 on a rough wall side. The spatially averaged shear stress profiles for all roughness pitch ratios can be represented as a liner function, and the slope of each linear function varied depending on the local skin friction coefficient. The profile of the spatially averaged Reynolds stresses takes a maximum value at the roughness pitch ratio of 8. The peak value of the dispersive stresses increases with the roughness pitch ratio.

Direct numerical simulations have been performed for turbulent heat and momentum transfer in a plane channel with streamwise rectangular grooves. The grooves are introduced by the immersed boundary method. The temperature is supposed to be a passive scalar, and the Prandtl number is set to unity. It is found that the simple streamwise grooves can enhance momentum and heat transfer in comparison to a smooth channel, leading to dissimilar heat transfer enhancement. For the fixed friction Reynolds number and the groove width and spacing, the dissimilarity between heat and momentum transfer is enhanced as the groove depth increases. The dissimilar heat transfer is associated with the appearance of large-scale intense spanwise rolls over the grooves.

Fluorescent luminous microsphere plastic particles Lumisis Marker® is used as tracers to visualize the turbulent water flow. And velocity field is calculated by tomo-PIV method to obtain the velocity gradient tensor. The vortex structures are identified by Q-criterion and the vortex sheet structures exist surrounding them. We also report the applicability of Lumisis® to air turbulent flow by UV laser illumination. The flow field is visualized by tow colors (red and green) and the mixing process is visualized for the first time.

The complexity or randomness of the relaminarization process of an accelerated turbulent boundary layer was analyzed by use of Kolmogorov complexity. Velocity data from a hot-wire anemometer was compressed using a compression program running on a Windows PC. From the compressed data, the approximated Kolmogorov complexity, named AK, and normalized compression distance, NCD was obtained. The values of AK of fluctuating velocity, time derivative of fluctuating velocity, and squared time derivative of fluctuating velocity, decreased toward downstream. Therefore, it was found that the relaminarization process can be properly represented. During the relaminarization process, the distribution of AK differs between streamwise and wall-normal fluctuating velocities, which can be explained by the difference of distribution of the fluctuating velocity themselves. On the other hand, some aspects are different from the fluctuating velocity themselves. AK shows different distributions among fluctuating velocity, time derivative of fluctuating velocity, and squared time derivative of fluctuating velocity. In particular, the turbulent structure of the relaminarization process could be clarified in more detail by comparing the AK of the squared time derivative of fluctuating velocity, with turbulent energy dissipation. From the NCD of streamwise fluctuating velocity whose reference position is within turbulent region, the turbulent structure appeared to turn into different one. Therefore, by changing the reference position of NCD, we were able to capture important changes that do not appear in the turbulent statistics.

In the present study, dielectric barrier discharge plasma actuators (DBD-PA) arrays were used to blow the near wall low-velocity fluid into the log-region, so as to generate artificial large-scale motions (ALSMs) in a turbulent boundary layer (TBL). The DBD-PA was embedded in the wall to avoid disturbing the flow. To investigate the effect of the ALSMs on the burst phenomenon, a multi-point hot-wire probe consisting of a hot-wire rake and a single I-type hot-wire probe was used to measure the fluctuation velocity in the log-region and buffer layer in the TBL. As a result, it was confirmed from the mean velocity distribution in the log-region that there exists artificial low-velocity region flanked on either side by high-velocity region. These low- or high- velocity regions were regarded as the negative ALSMs (nALSMs) and positive ALSMs (pALSMs), respectively. The bursting phenomena were detected by the Variable-Interval Time-Averaging (VITA) method from the fluctuation velocity measured in the buffer layer. It was found that the bursting frequency under pALSMs is smaller than that of nALSMs. Furthermore, from the pre-multiplied power spectra calculated from the fluctuation velocity measured by the single I-type hot-wire probe, it was found that the near-wall turbulent fluctuation energy in a large range of frequency was suppresed under the ALSMs, especially under the pALSMs. This could be considered as that the nALSMs have lifted the near-wall flow having a high turbulent intensity out of the near-wall region and the pALSMs have introduced the flow having a lower turbulent fluctuation energy from the outer region into the near-wall region, therefore suppressed the bursting phenomenon.

In the present study, a convolutional neural network model based on U-Net architecture is constructed to estimate a turbulent channel flow from measurements on the wall. By assuming that we can measure pressure and shear stresses in streamwise and spanwise directions on a wall, we estimate velocity and pressure at several planes parallel to the walls inside the flow. We have conducted direct numerical simulation of a turbulent channel flow at friction Reynolds number, Re_τ = 150, to build a dataset for the machine learning. We investigate the estimation performances obtained by the present U-Net model. It is found that the estimation accuracy of pressure and three velocity components is high near the wall but deteriorates rapidly away from the wall. In the log layer, the model shows poor estimation performance especially for velocity components. The model can estimate pressure better than velocity components in the whole domain.

The sweeping jet has been attracted attention as active flow control and mixing acceleration devices. Especially, the studies of the flow separation control techniques using the sweeping jet have been reported recently. However, most of these results are mainly contribution of the dynamic forces and the study on the flow mechanism produced by a sweeping jet has not been understood sufficiently. Especially, the vortex structure and flow mechanism formed by the interaction between the sweeping jet and the main flow have not been understood sufficiently. The purpose of the present study is to investigate the detailed vortex structure produced by the interaction between the sweeping jet from the Fluidic Oscillator and the main flow by a numerical simulation.

The effects of velocity fluctuations of upper flow on turbulent structure of atmospheric boundary layer were experimentally investigated. The atmospheric boundary layer was reproduced in a wind tunnel. In order to produce various velocity fluctuations in the upper flow, three kinds of turbulence grids were set to the entrance of the test section. The streamwise and spanwise velocities were simultaneously measured by using hot-wires anemometer. Turbulence statistics were estimated. The vertical profiles of Reynolds shear stress, Ruw, for M25 and M50 decrease as z/L increase, on the other hands, those for M75 increase. The ratio of the integral time scale, Λ/Λ₀, for M75 were higher than those for M25, M50 at z/L=0.67. It is suggested that the effects on the turbulent structure of the atmospheric boundary layer depends on the scale of velocity fluctuations of the upper flow.

Three-dimensional structures such as longitudinal vortex appear in the two-dimensional vortex street wake of a cylinder. These structures are very complex structures. It is important to understand the details of the three-dimensional structures in the two-dimensional vortex street wake. The purpose of this study is to construct the three-dimensional structures from layered cross-sectional images and to understand its structures. In the flow visualization, four cylinders with different diameter were used to change the Reynolds number. The 3D structure constructed in this study could be observed from any angles and the details of its structures were revealed.

A sphere with a uniaxial through-hole of 6mm with a sphere diameter of 25.4 mm and density of 1360 kg/m³ was launched vertically upward toward the free surface of calm water. The effect of initial angle of the uniaxial through-hole was investigated from the time the sphere passed through the calm water surface until it attained its maximum displacement position. When uniaxial through-hole of the sphere and the water surface are horizontal, a slight rotation occurs in the sphere that rises in the water. As the initial angle of the uniaxial through-hole increases, the sphere over the water surface deviates greatly from the central axis.

The laminar-turbulent transition of a mixing layer excited by oscillating flat plates at an exit of a two-dimensional nozzle was experimentally investigated. The mixing layer was formed between the jet issued from the nozzle and the surrounding quiescent fluid. The plates oscillated vertically in relation to the mean flow. Upper and lower flat plates oscillated symmetrically. The oscillation frequency, 5 Hz, was two orders of magnitude smaller than the fundamental frequency of the velocity fluctuation in the natural transition process. The amplitude of the respective plates was 0.5, 1.0, 1.5, 2.0 and 3.0 mm. Mean and fluctuating velocity components in the streamwise and normal directions were measured by hot-wire anemometers. We examine how this localized perturbation influences the turbulent transition of the mixing layer. The phase variation of the flow rate at the exit of the nozzle increased as the plate amplitude increased, though they were not in a proportional relationship. The mean velocity profiles skewed at the exit of the nozzle with the large amplitude, though the skewness disappeared downstream. The mean vorticity thickness was thick with large amplitude. The irregular fluctuation velocity with large amplitude increased earlier, and the turbulent transition was accelerated.

The aim of this study is to merge vortex rings in the initial region of a round jet. Air is issuing from a round nozzle into still air. The jet Reynolds number at the nozzle exit condition is set to from 1000 to 10000. The stream wise cross-section on the jet centerline was visualized by a planer laser Mie-scattering (PLMS) method and recorded by a high-speed camera. For the jets at Reynolds number when the merging of vortex rings occur in the natural transition, the merging of four vortex rings is achieved by applying an acoustic disturbance with the frequency of vortex formation. The merging process of the vortex rings is observed from the visualized images, and the spatial distribution of the merging of the vortex rings is created and compared with the Reynolds number.

To minimize sa fe distance from hydrogen stations, drag characteristics of jet impinging to perforated walls were investigated. The non-perforated wall and the perforated walls with five different opening rates were tested. The load cells supporting perforated walls were used to measure the drag. The drag was normalized by the non-perforated wall's drag. The drag characteristics are classified into three regions. The normalized drag is increased up to the normalized distance 200, almost the constant between 200 and 250, and increased again from 250.

The control mechanism for the rectangular jet using a plasma actuator with the burst frequency is investigated by the numerical simulation. When the plasma actuator is activated with the 55Hz burst frequency, the 10% duty rate and the antiphase, the full width at half maximum increases about 3 times compared to that of the without control case. The control effect might come from the two vortices which are created by the plasma actuator and the different vorticities for the vortices generated on the upper and lower surface of the rectangular jet. The first vortex from the upper surface of the rectangular jet is mixed the second vortex from the lower surface of that and is convected lower direction. The reverse phenomenon occurs for the second vortex from the upper surface of the rectangular jet. Therefore, the diffuse effect could be facilitated. On the other hands, when the condition for the plasma actuator is the 700Hz burst frequency, the 10% duty rate and the antiphase, the full width at half maximum decreases about 0.8 times compared to that of the without control case. The reason is that the vortex formed rapidly by the plasma actuator is smaller than the vortex from the mainstream which no longer generate afterwards the activation of the actuator, thus the diffuse effect could be suppressed.

This study is concerned with the flow control of backward-facing step flow by a dielectric barrier discharge （DBD） plasma actuator. The DBD plasma actuator is mounted on the step edge, and the DBD plasma actuator is driven by burst modulation. Flow visualization behind the step and pressure measurement on the wall were carried out. As a result, when the mainstream velocity is changed, the burst modulation frequency that moves the reattachment point position most upstream or downstream changes, but the effect of the duty ratio on the reattachment point position does not change. In the case that the reattachment point moves upstream, the position where the vortex generated due to the instability of the shear layer changes depending on the off-time length of the plasma actuator which calculated from duty ratio and burst modulation frequency. When the duty ratio becomes larger than a certain value, the effect of control is reduced. And the effect of the length of this OFF time on the reattachment point position does not change even under conditions where the mainstream velocity is different.

The purpose of this study is to improve the power generation efficiency of the self-powered turbine flowmeter developed by the authors. To increase the efficiency, the NACA airfoil was applied to the blade of the rotor. It was shown that the efficiency is favorably improved over the entire range of the flowrate by using the NACA airfoil blade.

Since a fan cooling for electronic devices is conventionally used, it is important to understand the flow phenomena to improve cooling performance. In this study, we investigate the heat-flow characteristics of around two small-axial fans under the influence of varied the distance between centers of two fans and the distance between the fans and the heating plate 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. As a result, we obtain the following findings: when the distance between the centers of the fans is large, a blow-up flow is formed due to the interference of the flows generated by each fan, and the heat transfer performance is low just below this blow-up flow; when the distance between the fan centers is narrow, no blow flow occurs, and the area with high heat transfer performance shows a large-scale asymmetrical distribution; when the distance between the fan and the heating plate is reduced, the peak value of the Nu number increases and the distribution of the peak value is distributed in a narrow range.

In this study, we evaluated the analysis algorithm and control window with standard deviation and non-detection vector rate to find the appropriate analysis conditions in PIV analysis of the interference fringe images by OFI (Oil Film Interferometry) adapted to the suction surface of a two-dimensional airfoil in high lift conditions with Reynolds number around 8×10⁴. As a result, X/C=0.75 where the flow was relatively simple showed a similar trend to the PIV analysis using particle images, and the analysis conditions were suitable for the control window of 32×32 pixels using the Direct Spatial Domain Correlation method or WIDIM. Near the reattachment point with secondary flow at X/C=0.17 to 0.25, the analysis conditions were suitable for the control window of 128×128 pixels using the Direct Spatial Domain Correlation method, and if higher resolution was required, the use of WIDIM to reduce the non-detection vector rate was suitable. The velocity vector data obtained from the PIV analysis was used in the OFI analysis to successfully measure the wall shear stress and friction velocity in areas with secondary flow.

The flow structure around a 1/2 scale Ahmed body was investigated in a circuit-type wind tunnel by mean of the flow vector sensor (FVS), the I-type hot-wire probe and the oil film method. The FVS is a thermal flow sensor of MEMS, which is able to measure the instantaneous vector of the surface flow. The drag force of the Ahmed body decreases suddenly around a slant angle of 30°. The change of the wake structure of Ahmed body from the three-dimensional separated flow to the quasi-axisymmetric separated flow can be measured. The flow direction of mean vectors by the FVS on the rear slanted surface of Ahmed body agrees with the flow pattern by the oil film method. It is found that the fluctuation of the wall surface flow measured by FVS on the rear slanted surface of Ahmed body is correlated with the velocity fluctuation measured by the I-type hot-wire probe in the its wake.

This study focused on the characteristics of a cavitating jet issuing from a small rectangular orifice with different aspect ratios. The orifice size is similar to the control port in a water hydraulic spool valve. The authors conducted the flow visualization to capture the overall and cross-sectional jet images. The maximum Reynolds number based on the equivalent diameter is approximately 1.3×10⁵. The results show that cavitation bubbles generate in the jet shear layer and just downstream the jet exit. The shape of the cavitation region varies with the streamwise position except for the two-dimensional orifice having an aspect ratio (AR) of 38.3. In the square orifice, the cavitation region rotates at 45 degrees from the jet exit like an axis switching. The cavitation region in the jet of AR=6.76 deforms from the rectangular shape. On the other hand, under the large aspect ratio condition, the shape of the cavitation region is maintained relatively downstream.

Jets are one of the most fundamental flows used in industry, and among them, impinging jets are used in industrial applications such as cooling of gas turbine blades and electronic devices. Because of the high heat and mass transfer coefficients near the stagnation point, impinging jets are used in various fields as a highly efficient means of heating, cooling, drying, and cleaning surfaces. On the other hand, the heat transfer performance of the impinging jet decreases rapidly away from the stagnation region, and it is necessary to increase the heat transfer area by eliminating the non-uniform heat transfer distribution. In previous studies, it has been shown that the bifurcation structure of the jet, which is characteristic of a blooming jet, can be obtained by combining inclination and rotation control of the jet with axial excitation, and that higher mixing performance can be obtained than that of the blooming jet. In this study, direct numerical simulations (DNS) of a inclined rotating impinging jets are conducted to evaluate the changes in flow characteristics and heat transfer performance. As a result, it was found that the controlled jet with inclined and rotational motion was strongly affected by the distance between jet and impinging wall.

In our previous research, we have been able to form branching flows corresponding to the number of excitation holes by adding synthetic jets to a round jet. We expect that a helical vortex structure is formed in the initial region of a round jet by adding a phase difference to the synthetic jets. Numerical simulations are carried out using an open source software OpenFOAM. It is confirmed that the helical vortex structure is produced in the initial region of the round jet. The formed helical-vortex-structure is visualized by the second invariant of velocity gradient tensor Q, in terms of the number of excitation holes, the amplitude, and the excitation frequency of the synthetic jets.

We have developed an experimental appratus that generates turbulent flow in a rectangular chamber by the interaction of a group of supersonic synthetic jets, each of which is generated by a piston/cylinder actuator driven by a motor. The four actuators are placed on two opposite sides of the chamber, and the supersonic jets are repeatedly ejected into the chamber from the jet holes on the top of the cylinder. The jets are ejected through either sharp-edged orifice or convergent-divergent nozzle. Shadowgraph visualization and particle image velocimetry (PIV) are used to investigate the characteristics of the generated turbulent flow field. The temperature increase due to the pistons causes the evaporation of engine oil inside the actuators. The PIV uses oil mist that is generated by condensation in the chamber when the hot jets with oil vapers are mixed with cold air. In the shadowgraph visualization, it was found that strong density fluctuations associated with small-scale vortices exist in the flow. For both orifices and convergent-divergent nozzles, the turbulent Reynolds number at the center of the chamber reaches about 1000. The relation between the turbulent Reynolds number and velocity derivative flatness is consistent with previous studies on incompressible turbulence. Although the supersonic jets are repeatedly ejected from two sides of the chamber, strong inhomogeneity due to the jets is not observed near the center of the chamber.

Side jets are ejections formed at the initial region of a jet. In recent experiments, it is confirmed that when the jet velocity is increased or decreased, a hysteresis phenomenon occurs in the side jets. That is, the value of jet velocity is different, when side jets are generated or disappear. It is expected that the phenomenon is related to the acceleration of jet velocity. Since an experiment for the effect of the acceleration on the side-jets formation is difficult, the aim of this study is to investigate the hysteresis phenomenon by a numerical simulation using open source software OpenFORM.

There are two kinds of seeds that fly by wind, pappose seeds and winged seeds. The former is known to make parachuting flights, which utilize the drag force acting on the pappi. Each flying seed is designed to have a small rate of descent and to float in the air current as long as possible, according to their environmental conditions. The dandelion which is a common example of wind-dispersed plant has an intricate haired pappus facilitating flight. In the present study, in order to investigate the effect of the posture of the pappus with respect to the flow direction on generating drag, drop tests were performed and the flight behavior of the pappus is observed by using a high speed video camera. The falling speed of the pappus is strongly dependent on posture in the air current.

The purpose of this research is to verify the effect of the shape of the upper and lower edges of the shield, designed by the authors to reduce the inflow of droplets into the face shield, using visualization experiments. The research also verifies the validity of the results of the LES analysis performed in the previous research. Two different face shield shapes are used for the verification: one with small plates at the top and bottom of the shield to guide the sneeze flow, including droplets, over the head and under the chin, thus physically inhibiting the flow entering into the shield. The other is a shape with a small plate and a brim at the top and bottom ends. The aim is to prevent the inflow of droplets arriving later by turning the sneeze flow forward, which reaches the shield surface before the droplets. The results indicate that the shape of the upper and lower edges of the shield can be improved to inhibit the generation of entraining flow near the upper and lower edges of the shield, and the amount of droplets entering the inside of the shield can be significantly reduced. These results are in good agreement with the results of the previous LES.

Passive control by changing the manifold shape is performed for multiple-passage channel flows. As for the manifold shape, the effects of manifold volume and corner curvature are investigated. The multiple-passage is a parallel flow type, and consists of the five branch channels. The channel flows are investigated from the view point of flow uniformity and pressure loss. Experiments are performed for the Reynolds number based on the bulk velocity and hydraulic diameter at the inlet channel from 6.0×10² to 2.0 ×10³. The wall static pressure is measured, and the pressure loss and flow rate in the branch channel are evaluated. The uniform flow distribution is realized by increasing the volume of the outlet manifold. The pressure loss is reduced by rounding the corners of the outlet manifold.

To improve the energy efficiency, it is important to increase the heat transfer coefficient per flow resistance. Previous studies have reported that the airfoil tube provides a better heat transfer coefficient per flow resistance than the circular tube. However, there are not many studies on heat transfer in the case of multiple configurations of airfoils. In this study, a comparison of the heat transfer efficiency of a staggered arrangement of airfoil tubes and circular tubes is performed using numerical analysis. Under the conditions of the present analysis, both the pump power and the total heat transfer of airfoil tubes are superior to those of circular tubes.

This study is directed toward the application of temperature fields due to pulsating flows in automobile exhaust pipes for active control. The temperature fields in multiple cross sections of pulsating and steady flows of high temperature air with an average Reynolds number of about 60,000 flowing in a horizontal straight tube with a rectangular cross section were measured with thermocouples. Measurements were made under pulsating conditions of 15Hz, 30Hz, 45Hz and 90Hz to investigate the change in heat transfer characteristics with pulsating frequency. As a result, it was found that the heat transfer was suppressed by the pulsating flow. At a pulsation frequency of 30 Hz, which is close to the resonant frequency of the piping, the heat transfer was suppressed more than in other pulsation conditions. In this condition, a strong reverse flow occurred. The strong reverse flow made it difficult for heat to escape, and heat transfer was suppressed.

A hemispheric head cyclone separator has higher collection perfoemance of fine particles compareing with a conventional cyclone separator with a cylindrical head. In order to optimaize the shape model with hemispharic head, we have investigated the air flow in various shape cyclone models by making numerical and experimantal approaches. In the near future, we will propose the compact cylone with hemisperical head for use as an intake side pre-filter for fuel cell vehicles. In this study, we forcus a dust holder which requrires the most large space to achieve space-saving of an installation space in vehicles. We have used direct numerical simulation (DNS) to investigate the internal airflow characteristics of four hemispheric head cyclone separators of different lengths with the diameter of the dust holder being 55% of that of the previous model (D-M). The cyclone separator flow is known to have a large upward vortex swinging around the core, and it has been reported that the cyclone separator flow affects the collection rate of fine particles in a cylindrical cyclone flow. Therefore, the minimum pressure position in each x-y section was detected every 0.01 s as the instantaneous vortex core position of the upward vortex, and the turning center position and radius of the swing were calculated from the trajectory of the vortex core position. The larger the radius of runout, the lower the frequency of runout tends to be. As a result, the difference in the effect of the runout motion of the upward vortex on the reduction of the collection rate is expected to be small in the model with a smaller diameter and varying length of the dust holder, but the model with a larger radius of runout at the bottom of the conical (SD-L2), where particles transported to the outside are more easily captured, is expected to have a larger percentage of reduction in collection due to runout.

This study was performed to numerically simulate two-liquid flows in a micro T-mixer with baffle. The baffle length and setting angle were varied to investigate their effect on the mixing performance. The mass conservation, alpha diffusivity and momentum conservation equations as the governing equations are used to mathematically model the problem, and a transient solver twoLiquidMixingFoam for mixing of 2 (two) incompressible fluids was used to solve the governing equations. As a result, the length and setting angle of the baffle shown a significant influence on the mixing performance, and a good mixing efficiency can be obtained for a baffle of length l_b=0.07mm at a setting angle of θ_b=30o and at a varying pitch.

In this study, a numerical analysis was performed to simulate two-liquid flow in a vortex micromixer under different flow condition. The micromixer is composed of a microchannel and a vortex chamber with 4 inlets. Each of the liquids is introduced through 2 alternative inlets, and then they are allowed to pre-mix in the vortex chamber before entering the channel. The length of the vortex chamber was changed to investigate the optimal mixing performance. As a result, the length of the vortex chamber shown a significant influence on the mixing performance, and the longer chamber obtained a higher mixing index.

We performed direct numerical simulation of subcritical transition in Taylor–Couette–Poiseuille flows (TCPf) with a high radius ratio of η = 0.883. TCPf is a combined-shear flow consist of the azimuthal rotating cylinders and the axial pressure gradient in a concentric annular pipe. Various wall-bounded shear flows were found to undergo spatiotemporal intermittency with localized-turbulent structures in the forms of spots or stripes. For instance, the turbulent spot and stripe seen in a counter-rotating Taylor–Couette flow (TCf) are known as intermittent turbulent spot (INT) and spiral turbulence. On the other hand, the turbulent stripe in an annular Poiseuille flow (aPf) exhibits helical turbulence with a high radius ratio of η ≥ 0.5. The control parameters are the Reynolds numbers Re_in and Re_out based on the cylinders’ rotating velocities and the axial pressure gradient function F(P). In this study, we chose the Reynolds numbers at which INT occurs for the pure TCf, and increased F(P) gradually until the Poiseuille component was dominant. The INT with interpenetrating spirals in the F(P) = 0 has changed to complete laminar flow via forming spiral turbulence. The flow fields exhibit turbulent stripes again in higher F(P) than the laminar conditions (Re_τ,z ≳ 40). This turbulent stripe may be helical turbulence since the Poiseuille component is more dominant than the Couette component in this stage.

It was known that the circular pipe flow with a large initial turbulence was stable up to Reynolds number 1500, and became unstable over that, but did not transition to turbulence up to Reynolds number 2300. CFD by the central finite difference method was applied to the circular pipe flow with a large initial turbulence in the Reynolds number range over 1500. The calculation result of the unstable flow showed that the pressure value at the center was smaller than that at the wall surface. As a result of the experiment, it was found that the static pressure at the center was larger than that at the wall surface, contrary to the calculation result. It was considered that the transient Reynolds stress acted like pressure, and it was confirmed that the calculated value of the sum of the transient Reynolds stress and the pressure was larger at the center than at the wall surface. Furthermore, when the transient Reynolds stress was added to the right side of the Navier-Stokes equation of motion, it was confirmed that transient Reynolds stress had the effect of delaying the divergence of solution. The effect was remarkable when the Reynolds number was 1500 to 2000. In an actual circular pipe flow with a large initial turbulence, the fact that the flow could maintain a laminar flow up to Reynolds number 2300 was considered to be due to the transient Reynolds stress.

I calculated the laminar oil flow in a circular pipe by using OpenFOAM. The pressure loss coefficient of the flow was nearly 64/Re.I also calculated the flow through orifice plates with various diameter ratios.I found that the flow coefficient of the orifis is affected by the diameter ratios and temperature change.

Visualizations of transition process triggered by disturbance injection in pipe flows were conducted. In order to investigate the influence of hairpin vortices formed by interaction between the mainstream and the injected jet on the transition process, the hairpin vortices and the secondary turbulence including turbulent puffs are distinguished from each other by the fluorescent dye and flake-like particles. Timeline images on the visualizations indicated different creation processes of a puff depending on the conditions, the disturbance amplitude A and disturbance injection time L*. In addition to this, a device generating microbubbles in the pipe by electrolysis of water was installed upstream the jet injection point, and the influence of generating microbubbles on the transition process was visualized. The results indicate that the microbubbles modify radial position of the hairpin vortex street and this modification enhances the transition.

We performed a parametric study of high-aspect-ratio channel flows with spanwise side-walls, i.e., very thin and wide duct flows, in subcritical turbulent transition using direct numerical simulation. We investigated the finiteness in a spanwise domain so-called side-wall effects of a duct flow on the localized turbulence and verified a relevance of the transition process and directed percolation (DP) universality. Many shear flows could sustain turbulence globally with large scale intermittent structures of coexisting laminar and turbulence regions for very low Reynolds number which is enough low from the upper critical values. Also, a transition of some flows can be considered as DP universality class. A duct flow with very wide spanwise domain forms spatial-temporal intermittencies such as the form of stripe or band shown in the channel flow near the lower critical Reynold number just before laminar flow. The present duct flows sustained the intermittencies for Re > 1000, whereas the channel flow formed isolated turbulent bands stably near the global critical Reynolds number as low as Re ≈ 700. In the present widest duct flow, a turbulent band collides on the side-walls and reflects or reverse travels for Re ≈ 1000. Below the global critical Reynold number, a frequency of a turbulence band collapsing goes extremely high. Thus, the spanwise finiteness inhibits the isolated turbulent bands sustaining and prevents deviated from two-dimensional DP transition for 700 < Re < 1000. As a result, the duct flow transition converges to (2+1)-D DP with Re_g ≈ 1000 by increasing the distance between the side-walls.

Our country depends on import for much of energy and mineral resources. From these background, marine resources lying in territorial waters and exclusive economic zones are attracting attention. Underwater robots are mainly used to search for marine resources. Conventional robots with screw type thruster have various problems, and in order to solve them, fish imitation robots are being developed. In recent years, in order to help its efficient locomotion, we try to apply a wing with sinusoidal leading edge, which is a biomimetic of humpback whale, to the pectoral fin part of fish imitation robot. In this study, to clarify the suitability of the wing with sinusoidal leading edge for the fish imitation robots, we investigated the dynamic lift acting on the pitching wing under the condition of two kinds of offset angles, α₀=5°and α₀ =15°. The results were compared with those in quasi-steady case and in straight leading edge case to clarify the effect of pitching motion and sinusoidal leading edge. As a results, it was found that at α₀=5°, the lift force of pitching wing is almost the same as the quasi-steady state case independent of leading edge shape. However, at α₀=15°, where the stall occurs, wing with sinusoidal leading edge can maintain higher lift force than the straight leading edge case.

Since penguins spend long time in the sea, their body surface presumably evolved to reduce fluid friction drag. The barbs of the body feathers are oriented in the longitudinal direction and form a group of microgrooves, so we hypothesized that they work as riblets to reduce friction drag. In addition, considering various maneuvers of penguins, the drag reduction effect of the barbs is desired to be robust to changes in the yaw angle to the flow direction. To test those hypotheses, we fabricated trapezoidal riblets based on the morphology of the barbs, followed by drag measurement of flat plates with the riblets in a water tunnel. We measured dimensions of the real barbs of penguins and designed the riblets based on the spacing and cross-sectional shape of them. We fabricated two types of riblets on polyimide films by laser scanning ablation. One was an actual-scale riblet with yaw angles of 0° 15°, 30°, or 45° which was same size as the barbs. The other was a small-scale riblet to reproduce the surface flow condition of a swimming penguin in our water tunnel. As a result, the actual-scale riblet showed the largest drag reduction ratio of 1.26% at a yaw angle of 15°. This indicates that the drag reduction effect is robust to yaw angle up to 15°. The small-scale riblet showed the largest drag reduction of 1.97% when s⁺ was 1.59. The drag reduction effect was also observed in all s⁺ ranges, indicating that the barbs of penguins have a drag reduction effect during swimming. These results indicate that penguin barbs have the drag reduction effect in the all range of swimming speeds, and that the drag reduction effect is greatest when the yaw angle is 15°.

Drag reduction effect of riblet in turbulent Taylor-Couette flow is experimentally investigated. A motor and a torque meter are installed at the top of the cylinders to investigate the effect of drag reduction by torque measurement. The outer cylinder is fixed and the inner cylinder rotates on its cylindrical axis. Two inner cylinders manufactured by 3D printer, one with a flat surface and the other with sinusoidal riblet, were compared to investigate the drag reduction effect. As a result of torque measurement, it was confirmed that the drag increased about 20% at maximum by sinusoidal riblet. In order to achieve high drag reduction effect, it is necessary to investigate riblet that focuses on Taylor vortex in addition to turbulent components.

Numerical simulations of fully developed turbulent channel flows with riblet surfaces were carried out at the friction Reynolds number of 180. Influence of particles adhering on the blade-type riblets on the drag reduction effect is investigated. A one-way coupling method is employed: the particles are affected by the fluid, but the fluid is not affected by the particles. The simulation consists of three steps: first, calculation of the adhesion position of the particles to the riblets in the turbulent channel flow; second, the riblets shapes are changed based on the particle adhesion; third, calculation of the turbulent flow over the particle adhered riblet. As a result, most of the particles adhered to the upper wall and the top of the side wall of the riblets, while number of the particles adhering to the vicinity of the channel wall is very small. Due to the particle adhesion on the riblet, the drag increases as compared with the clean riblet case and the plane channel flow. We confirmed that the Reynolds shear stress profile is to be consistent with the drag variation.