A finite difference method (FDM) is easier and cheaper to use than a spectral method (SM) in flows with complicated geometry, but is inferior to a SM in both accuracy and resolution. In this paper, effects of both accuracy and resolution in a FDM on the quality of incompressible turbulent flow description were studied by comparing a SM with a FDM in direct numerical simulations (DNS) of the three-dimensional freely decaying homogeneous turbulence with 128³ and 256³ grids. We adopted the higher-order method of lines and Fourier spectral method as a FDM and a SM, respectively. The value of the kinematic viscosity was determined so that the minimum value of K_maxη in time was about 2 in the SM with 256³ grids. Here, K_max and η are the maximum resolved wavenumber and the Kolmogorov length scale, respectively. The statistical quantities, such as the total energy, the energy dissipation ratio, obtained by the higher-order method of lines were in good agreement with those by the SM. Furthermore, comparing their DNS fields, we found that the performance of the 4th-order method of lines with 256³ grids was comparable to that of the SM with 128³ grids in which the minimum value of K_maxη in time was about 1.

View full abstract

A lifting-line theory of a wing in uniform shear flow is presented and also a critical shear parameter where aerodynamic characteristics abruptly change is found by obtaining analytical expressions of the characteristics. The governing equation with respect to potential function defined by Karman is derived and reduced to two ordinary differential equations by separation of variables. A general solution is obtained by linearly combining the fundamental solutions of the ordinary differential equations. An induced attack angle is derived from the general soultion and a condition of far downstream. A relation of a lift force to an effective attack angle is described by the Taylor series expansion in a vicinity of geometrical attack angle. From the induced attack angle and the lift force, a lifting-line equation is obtained, and solved by Gaussian elimination method. Effects of shear parameter and aspect ratio on the local lift, total lift and induced drag coefficients are clarified through some numerical calculations.

View full abstract

Hydraulic transients such as water hammer or column separation are very important phenomena to be considered in the design of pump, pipeline transport system and so on. Majority of the numerical analysis of these phenomena has been performed with 1-dimensional analysis, but it is hard to understand the internal phenomena such as wave reflection caused by an existence of the pipe wall and variation of pulsating flow condition. In addition, modeling of the water column separation process is not so easy. In this study, in order to clarify the hydraulic transients in 2-dimensional channel, a numerical analysis is carried out by solving the governing equations taken account of the compressibility of the gas-liquid two-phase media.

View full abstract

A study was conducted to obtain fundamental flow characteristics for a spool valve whose inlet and outlet port position was set at 90° angle. Basic experiment was carried out to obtain basic flow and pressure relationship, which was used to the basis of the three dimentional computer simulation. By examining experimental data with the numerical results, the spool configuration was connected in order to give the consistency between the experiment and numerical analysis. Various flow characteristics were verified by the numerical simulation, and detailed flow field and their associated flow parameters of the spool valve were obtained in the present study.

View full abstract

Wake properties of a rectangular plate with aspect ratio AR of 2 and 3 are investigated experimentally to identify statistically. It is characteristic of the near wake of a rectangular plate that mean velocity profile has two inflection points on the major-axis. One of the inflection points is yielded by the vortex shedding and the other is located within the recirculation region. The growth rate of a half width on the minor-axis is greater than that of the axisymmetric bluff bodies, such as a sphere and a circular disk, it follows the growth rate of a half width of the two-dimensional wake. The reduction rate of the maximum velocity defect is also greater than wakes of axisymmetric and two-dimensional bluff bodies. These wake properties of a rectangular plate are similar with that of an elliptic wake with the same AR. However, the existence of the corner of a rectangular plate, which is the geometrical difference between the rectangular and elliptic plate, only affected the maximum turbulent intensity on major axis.

View full abstract

In this study, suppression of fluid forces, particularly fluctuating fluid forces, acting on a circular cylinder in uniform flow was investigated. Experiments were carried out with the Reynolds numbers fixed at 5.5×10⁴, which is in the range of subcritical Reynolds numbers. Fluid forces acting on the circular cylinder was suppressed by setting two tripping rods in the neighborhood of the surface of the circular cylinder to control the flow over the surface of the cylinder. The angles at which the tripping rods were set were changed within the range of α=+20∼±60° from the stagnation point, and the ratio of tripping rod of diameter d to circular cylinder diameter D was changed within the range of d/D=0.08∼0.12. It was found that the largest reduction in fluid forces occurred when the tripping rods were set at angles in the range of α=30∼40°. In this condition, the time-averaged and fluctuating drag coefficients, C_D and C_Df, were decreased down to 73% and 64% respectively, and the fluctuating lift coefficients C_Lf, was decreased down to 87%. The generation of fluctuating lift associated with vortex shedding from the cylinder was suppressed almost completely in this condition.

View full abstract

A pure tone noise generated from a multitube once-through boiler (test product) at non-combustion testing was investigated in order to clarify the generation mechanism and the oscillating frequency. The acoustic analysis of a boiler, the water tank test and the two-dimensional CFD analysis for an in-line tube array model were conducted. As a result, it was found that the vortex in wake of the tube shed periodically with the same frequency as the acoustic natural frequency of the boiler, namely the aerodynamic self-sustained tone was generated. Moreover, a wind tunnel test of the in-line tube array model with the various diameters and intervals of tubes was conducted. As a result, it was found that the values of Strouhal number St based on the length of separation shear layer were about 0.3∼0.6, and the phenomena of periodic vortex shedding on a boiler were similar to those on a bellows flow and a cavity flow. This paper presents some useful experimental data for low noise design of a boiler.

View full abstract

Effect of periodic perturbation was investigated experimentally on the management of a stronger wall jet with a streamwise vortex. A half-delta wing located on the wall generates the streamwise vorticity and the periodic perturbation is added by varying in angle of attack of the wing. Measurements of mean velocity, secondary velocity vectors, and streamwise vorticity were examined to see effect of the periodic perturbation on the evolution process of the wall jet and streamwise vortex. The periodic perturbation accelerates development of vortex radius in spanwise direction over the initial stage and increases relative strength of spanwise secondary velocity vector to the transverse velocity vector. The streamwise momentum flux is increased by interaction between streamwise vortex and wall jet. However, the periodic perturbation reduces increasing rate of the momentum flux.

View full abstract

In order to reduce pressure loss and asymmetry, a flow control device was applied to the complex flow field in a 90-degree bend pipe : a trip wire was attached on the inside wall of the curve at appropriate upstream position to the 90-degree bend. The trip device spans half circumferential length, that is, it is called as "half-trip" in the present experiment. To see ability of the device to control the complex flow, pressure loss, axial mean velocity profile and secondary current were measured for Reynolds number range Re=50000-82000. The experiment was made with relatively small rate of curvature to pipe radius 2R/D=1.43. Wall static pressure measurements certainly show that the half-trip is effective to reduce pressure loss as well as asymmetry of the axial mean velocity at exit of the bend pipe. The maximum reduction rates of pressure loss coefficient is 23% for trip device height k=3.0mm and 5.0mm. A possible mechanism is discussed with available experimental evidences of axial velocity, secondary velocity vectors and oil film pattern observations.

View full abstract

The blade-vortex-interaction (BVI) of a helicopter's two-blade rotor is analyzed by the three-dimensional discrete vortex method for various conditions including the helicopter's body. The rotor is represented by source panels and the decent vortices are shed from the tips and the trailing edges of the blades as vertex rings and a part of each vortex ring is kept within the blades. The body is also expressed by source panels. It is found that the thrust fluctuation on the blade occurs rather randomly in the region of the azimuth angle 0 to 180 degrees. The fluctuation is stronger in descending case than in horizontally flying case, and the fluctuation occurs even in the region of azimuth angle larger than 320 degrees in the descending case. The helicopter's body does not affect the characteristics of the rotor aerodynamics, but the disturbance in the wake slightly glows. It is also shown that combinations of the vortices is effective for decreasing the number of vortices without affecting the computing results.

View full abstract

A Wells turbine for wave power conversion has hysteretic characteristics in a reciprocating flow. The counterclockwise hysteretic loop of the Wells turbine is opposite to the clockwise one of the well-known dynamic stall of an airfoil. In this paper, the mechanism of the hysteretic behavior was elucidated by an unsteady 3-dimensional Navier-Stokes numerical simulation. It was found that the hysteretic behavior was associated with a streamwise vortical flow appearing near the blade suction surface. In the accelerating process of axial flow velocity, the vortex is intensified to enlarge the flow separation area on the blade suction surface. In the decelerating flow process, the flow separation area is reduced because of the weakened vortex. Therefore, the aerodynamic performance in the accelerating flow process is lower than in the decelerating flow process, unlike the dynamic stall. Based on the vortex theorem, the mechanism to vary the intensity of the vortex is explained by the trailing vortices associated with change in the blade circulation.

View full abstract

This study deals with the evanescent wave Doppler velocimetry, which has been proposed by one of the authors for measuring flow velocity in a wall's near field. In order to evaluate the applicability of the velocimetry to a viscous flow, the velocity measurement in a Couette flow of water between a gap of 1.5mm, which is considered to have a linear velocity distribution, has been performed. Although the measured velocities are slightly smaller than the corresponding ones that are predicted assuming the linear velocity distribution, the applicability of the velocimetry to a viscous flow has been evaluated. This study also shows tracer particles that are suitable for the velocimetry.

View full abstract

In highly rarefied gas flows, there appear nonequilibrium phenomena not only between translational and rotational energy modes (non-equipartition) but also deviation of rotational energy distribution from the Boltzmann distribution. To analyze these highly rarefied gas flows, we established the experimental system for 2R+2 N₂-REMPI and applied it to the measurement of rotational temperature in a supersonic free molecular flow of nitrogen. In this study, a strong nonequilibrium in rotational energy mode has been found in a nitrogen free molecular flow. The data in the Boltzmann plot, obtained from the measured REMPI spectra, cannot be fitted by one line but approximately by two lines, revealing the non-Boltzmann distribution of rotational energy in the ground state. The mechanism of the deviation from the Boltzmann distribution is also discussed using the state-to-state transition rates for collisional relaxation.

View full abstract

We propose a combination scheme for the lattice-Boltzmann method (LBM) to reduce the unnatural distortion, which is caused by the crossing angle between lattice and boundary and appears in numerical solutions at high Reynolds number flows. The suitable ratio for blending 15- and 19-velocity models was derived theoretically to be 1 : 2, to eliminate the error in calculation of three-dimensional, viscous incompressible flows. Then we applied it to one of the most critical case ; namely, a fully developed turbulent flow in a straight pipe of circular cross section was directly calculated by the cubic grid. The mean velocity profiles obtained by 15- and 19-velocity models were significantly distorted and they had opposite properties. Our mixed model, however, successfully reproduced the axisymmetric profile of time-averaged flow field and showed reasonable agreement with available database.

View full abstract

The effects of thermal stratifications on turbulent mixing and chemical reaction in the liquid mixing layer were investigated by both laboratory experiments and numerical simulations. The experiments were conducted in neutrally, unstably and stably stratified water flows with a rapid chemical reaction. Instantaneous velocities and concentration were simultaneously measured using a combined technique with a two component laser-Doppler velocimeter (LDV) and a later-induced fluorescence (LIF) method. Turbulence quantities such as the mean concentration of chemical product and turbulent mass fluxes were estimated and the effects of thermal stratifications on the quantities were clarified. Further, the large-eddy simulation (LES) using the large-eddy probability-density function (LEPDF) model as a SGS model was applied to the thermally-stratified reacting liquid flows. The results show that the effects of the stratifications on the turbulent mixing and chemical reaction can be well explained by the LES.

View full abstract

Experiments have been performed on turbulent channel flow with injection through an inclined slit. The inclined angle is changed 30°, 60° and 90° at the main stream velocity 10m/s. The dimensionless injection flow rate is varied three steps in ranging 0.067∼0.200. The cooling effectiveness, wall static pressure, wall shear stress, velocity profile and turbulent intensity are measured behind the slit. The cooling effectiveness increases with increasing the dimensionless injection flow rate. When the inclined angle is 30°, the effect of injection remains over long distance in the near-wall region, and the highest performance is obtained.

View full abstract

The turbulence structure of the liquid-solid vertical jet flows impinging on a flat plate has been experimentally investigated. The diameter of the particles is nearly ten times as large as the Kolmogorov length scale and the volumetric concentration of the dispersed phase was set at 0.13%. The particle speed is about 45% of the liquid velocity at the nozzle outlet. As a result of the measurements with LDA and PIV systems, it was shown that the turbulent intensity in the carrier fluid has been enhanced due to the effect of the large-scale particles. The increase of power spectra and dissipation spectra has been clearly found in the high wavenumber range, which corresponds to the particle length scale. Since the energy variation evaluated by the model, in which turbulence modification is caused by particle wake eddies, is equivalent to the measured values, it can be concluded that the turbulence enhancement is mainly caused by the wakes behind the particles.

View full abstract

Flow header to distribute a fluid to small multiple pipes is commonly used in boilers and heat exchangers. The system contributes to raise the heat transfer efficiency in the components. The flow distribution mechanism of the header has been studied and the calculation procedure for the design has been also recommended for a single phase flow. The most parts of these studies have been focused on the header of round pipes. However, thin rectangular flow header is preferable for compact design of heat exchanger. The experimental study was conducted with several kinds of thin rectangular and small square headers, and prediction method for distribution behavior was modified to take account of the frictional pressure loss in the header.

View full abstract

The authors have presented an active control method of surge occurring in a centrifugal compressor, using the control valve which is provided separately from the flow control valve. The surge control valve is operated with the plenum pressure as the feedback signal. The experimental results using the proportional control and the delayed feedback control have shown that the present methods are effective for the active control of surge.

View full abstract

A numerical scheme based on the control volume method to solve conjugate heat transfer problems in a system consisting of porous media and bounding walls has been developed using a two-energy equation model for porous insert. Appropriate wall heat flux expressions for the fluid phase and solid phase were obtained from the proposed heat transfer model at the interface between porous media and solid wall. The heat transfer model clarifies physics of the known thermal boundary condition at the interface. Predicted interface temperature distribution and Nusselt number show very good agreement with the available experimental data for the case of using metal foams as the porous insert. Effect of wall thickness on heat transfer to the porous media has been investigated and the importance of considering heat conduction in the wall as well as heat transfer in porous media was demonstrated.

View full abstract

Experiments were perfomed to study the combined effects of the height and thickness of micro-pin-fin on boiling heat transfer from silicon chips immersed in a pool of degassed or gas-dissolved FC-72. Six kinds of square micro-pin-fins with the dimensions of 30×60, 30×120, 30×200, 50×60, 50×200 and 50×270μm² (thickness t×height h) were fabricated on the surface of a square silicon chip with the dimensions of 10×10×0.5mm³ by use of the dry etching technique. The fin pitch was twice the fin thickness. The experiments were conducted at the liquid subcooling ΔT_sub of 0, 3, 25 and 45K under the atmospheric condition. The results were compared with previous results for a smooth chip and three chips with enhanced surfaces. The micro-pin-finned chips showed a considerable heat transfer enhancement in the nucleate boiling region and increase in the critical heat flux q_CHF as compared to the smooth chip. The wall temperature at the CHF point was always less than the maximum allowable temperature for LSI chips (=85°C). For a fixed value of t, q_CHF increased monotonically with increasing h. The increase was more significant for larger t. The q_CHF increased almost linearly with increasing ΔT_sub. The maximum value of allowable heat flux q_max (=84.5W/cm²), 4.2 times as large as that for the smooth chip, was obtained by the chip with h=270μm and t=50 μm at ΔT_sub=45K.

View full abstract

Numerical studies are carried out on the critical heat flux of ice accretion along a horizontal fine wire immersed in cold air stream with water spray. The parameters selects for the investigation are the mean diameter of the droplet, the wire diameter, and the ratio of the former and the latter. The temperature and velocity of a droplet that arrives at a wire are calculated, and based on the results the critical heat flux is estimated using the heat transfer equations previously proposed by other researchers. It is assumed that the droplet diameter distribution is the same as that of Rosin-Rammler. Finally, the characteristics of the critical heat flux are examined by comparing experimental data obtained by the present authors. It is found that the present simulation can successfully explain the experimental results, although it is needed to be modified for wider conditions.

View full abstract

Since fossil fuel like petroleum, coal, and natural gases will sooner or later run out in this or next century, natural energy like solar, wind, and biomass will rise as one of alterative energy sources. Most of the conventional solar thermal electric systems utilize high-temperature difference. In the present paper, an advanced solar Rankine cycle system with CPC (Compound Parabolic Concentrator) collector and steam accumulator of phase-change type is newly proposed. The phase-change steam accumulator with high temperature phase change material (PCM) is designed and evaluated both experimentally and theoretically. And this paper reports experimental results of the proposed system under actual solar radiation and under heat stored accumulator condition.

View full abstract

The fuel cell is developed as a power source which is friendly with the environment. To thin the fuel cell is useful to compactness and cost reduction. So, in this study, the fuel cell with thin separators of 0.6mm thickness including microchannels of 0.2mm was fabricated and experimented. Experimental results showed that the hydrophobic treatment of separators is indispensable to drain easily water condensed in the microchannels. Besides, as water produced at the cathode increases with current densities and fills some microchannels and the electrode pores the pressure loss at the cathode gas increases and the performance of the fuel cell decreases due to the reduction of gas diffusivity. However if the humidity of cathode inlet gas is controlled to vaporize produced water corresponding to current densities the thin separator with microchannels of 0.2mm thick could be used without any influence on the performance.

View full abstract

In this paper, we tried to make clear the effects of spray characteristics, especially spatial inhomogeneity of fuel droplets, on ignition probability of fuel spray. A mono-dispersed methanol spray was ignited by a discharge spark ignition system, and a large number of ignition tests were carried out. As a result, ignitability of fuel spray is affected not only by the instantaneous spray concentration just before ignited, but also by other probabilistic factors based on an inhomogeneity in spray. The ignition process can be classified into two processes, i.e. flame generation and flame propagation. The flame propagation process becomes major control factor of the result, as the discharged energy is sufficiently large. Further, the propagation process is strongly affected by inhomogeneity. Therefore, inhomogeneity of spray was expressed quantitatively by the inhomogeneity index and the relation between inhomogeneity of spray and ignitability was discussed. Results suggest that there is the optimum intensity and the scale of inhomogeneity for ignition. In this research condition, the characteristic scale of inhomogeneity was around 15mm for successful ignition and the both of the characteristic scale and its intensity of successful case were smaller (or less) than those of failed one.

View full abstract

Numerical computations of the inverse diffusion flame were conducted to investigate the flame structure at the upstream end of the diffusion flames adjacent to the burner rim and to make clear the mechanistic difference of the flame lifting between inverse and normal diffusion flames. The results obtained are as follows. (1) The upstream end of the diffusion flame is located at the air side near the nozzle rim and consequently the air velocity instead of the fuel velocity is sensitive to the flame lifting. This fact explains the characteristic difference of the flame lifting of the inverse and normal diffusion flames. (2) The flame structure of the upstream end of the diffusion flame consists of the diffusion flame structure in the radial direction together with the premixed flame structure in the axial direction. (3) It is estimated that the burning velocity of the premixed flame formed at the upstream end of the diffusion flame is considerably low as compared with that of the one-dimensional premixed flame because of the heat loss induced in the radial direction and the flame holding is attributed to the balance of the fluid velocity and burning velocity.

View full abstract

The quenching diameters of propane/air and methane/air flames were measured for propagation through a short circular hole, and the flame behavior was photographed using a highspeed digital video camera. As the flame quenching showed stochastic, two critical diameters were defined ; the critical quenching diameter, below which flames are always quenched, and the critical propagation diameter, above which flames always propagate through the holes. The equivalence ratio at which minimum quenching diameter of propane/air flame was obtained was much richer than the data obtained in the previous studies. This is considered to be the effects of Lewis number of unburned mixture and preferential diffusion of those species. The flame surfaces are convex toward the inlet of the hole. The flames showed complicated behavior due to buoyancy at the entrance of the hole. A flame let could be observed just above the inlet after the flame extinction on the surface of the glass. Almost all quenched flames extinguished above the inlet of the hole.

View full abstract

Applicability of a coal combustion model was validated by using experimental data of utility boilers. The coal combustion model had gasification and NO_X reaction submodels and it was developed by using a small drop-tube-furnace (coal feed rate 0.6kg/h). A turbulence combustion simulation program was developed by introducing the coal combustion model. The simulation program was validated by comparing with twenty three experimental results that contained different plant, load and coal data. The temperature difference between simulated and experimental results was less than 30 degrees C at the exit of the furnace. The characteristic of coal burnout that was decreased with increasing load was predicted. The NO_X concentration difference between simulated and experimental results was less than 15 percent. The coal combustion model was applicable to the utility boilers as same as the drop-tube-furnace.

View full abstract

In this study, the chemical kinetic analysis with a detailed hydrogen reaction model was conducted to examine the effects of temperature, pressure, and concentrations of fuel, oxygen and inert gas in Ar-H₂-O₂ mixtures on the auto-ignition delays. The results shows that the mixture temperature has a strong effect on auto-ignition delays, and that the richer mixtures ignite earlier because the increase in hydrogen concentration causes the decrease in the dilution of the inert gas in the mixture. Moreover, on the assumption that a mixture is composed of hydrogen, directly injected into a combustion chamber, and an ambient Ar-O₂ mixture, it is clarified that the inert gas concentration and temperature vary with the change in the hydrogen concentration in the mixture, and it can be supposed that the lean mixtures ignite earlier because temperatures of them are higher than those of the richer mixtures. And it is also supposed that the auto-ignition delays are shortened due to the promotion of the mixture formation.

View full abstract

Oxidation of dimethyl ether (DME) has been studied in motored internal combustion engine compression cycles. By applying a quadrupole mass spectrometer at reduced ionization energies to monitor the exhaust gas composition, it was found that fuel consumption is limited at relatively constant extent in certain ranges of intake gas temperature and fuel-air ratio of the low temperature oxidation (LTO) regime, which is in good agreement with simulations using Curran's kinetic mechanism. It is suggested that inhibition of chain branching by piled up HCHO is responsible for this behavior. The features appeared in experiment and calculation can be explained by simplified LTO of DME and HCHO oxidation.

View full abstract