This paper deals with the dynamics of a cluster of bubbles in a potential flow. The governing equations of the bubble cluster are derived by means of a series expansion of spherical harmonics. The three-dimensional translational motion and deformation, which are induced by bubble-bubble interactions and the pressure gradient due to fixed simple sources, are taken into consideration. Accuracy of the present theory is confirmed by comparison with the boundary element method. The theory is applied to the dynamics of traveling bubbles around two kinds of axisymmetric body. It is shown that the bubble deformation is greatly affected by the pressure gradient around a body. It is also shown that the bubble growth is reduced by bubble-bubble/bubble-wall interactions. The growth rate of interacting bubbles depends on both initial radius and initial location of bubbles. When one bubble grows much faster than the other bubbles, the growth rate of the other bubbles becomes much lower than that predicted by a single-bubble theory.

View full abstract

Energy loss of gas-liquid two-phase flow in a concentric annulus with a rotating inner cylinder has been experimentally investigated. The energy loss is closely related to the flow pattern and becomes almost constant in the spiral and ring-form flows observed under relatively high rotational speed. The ratio of the kinematic energy loss to the potential one in such flow regimes is well correlated by a linear function of the volumetric flux ratio of two phases. Drift-flux model and the two-phase friction multiplier of Chisholm are successfully applied to the estimation of the energy loss in this complex flow field.

View full abstract

Experimental investigations and numerical analysis were carried out to clarify the behavior of an interactive flow around two circular cylinders with different diameters, in particular, the phenomenon that the shear layer separated from the main cylinder reattaches to the rear surface due to the Coanda effect. Numerical analysis was simplified by applying a vortex method and a boundary element method. The results obtained were as follows. The reattachment due to the Coanda effect appeared intermittently and could simulated. Negative lift acted on the main cylinder, and the calculated lift and drag coefficient showed good agreement with those of experiments. The calculated Strouhal numbers were 20∼30% higher than the experimental ones.

View full abstract

In uniform flow, periodic Karman vortex shedding is observed in the wake of a stationary circular cylinder. When the cylinder oscillates transversely at a frequency close to the Karman vortex frequency, the vortex shedding is synchronized with the vibration of the cylinder. This is well known as the lock-in phenomenon in wind engineering. In this study, incompressible flow past a forced oscillating circular cylinder at Reynolds number 1100 is computed by finite-difference method using the moving grid technique, and the lock-in phenomenon is simulated by two- and three-dimensional calculation. In the results of the three-dimensional calculation, the lock-in region and drastic change of the phase angle between lift force and cylinder displacement show good agreement with experimental values. Furthermore, it is qualitatively shown that the correlation length of vortex shedding increases in the lock-in region, However, in the case of the two-dimensional calculation, lock-in phenomenon cannot be simulated with sufficient accuracy.

View full abstract

A new method of unstructured mesh generation is presented. The advancing front method and Delauney triangulation have been commonly applied to it. In the former method, interior mesh nodes are generated based on the nodes on boundaries and parameters defined by the user. Unsuitable setting of parameters, however, may lead to failure in mesh generation. On the other hand, the latter method can accomplish mesh generation without failure, though the user has to set all the nodes. The present method, based on Delauney triangulation, requires the user to set only a few nodes on boundaries, and the other mesh nodes are generated automatically according to the curvature or proximity of boundaries. The perfomance of the presented method is illustrated in a region with multiple bodies.

View full abstract

Recently, a two-phase thermal control system has been proposed to increasing the output power of large artificial satellites and space stations. For such a system, it will be important to establish a phase separation technique of a gas-liquid two-phase flow. In this paper, an impacting T-junction is proposed as the phase separation technique under microgravity. Ground experiments were first performed to clarify the phase separation characteristics. From the experiments, it was found that the inlet flow pattern and liquid velocity affected the phase separation most and that phase separation for stratified and wavy flow regions was difficult. Moreover, to verify the applicability of the T-junction under microgravity, experiments utilizing parabolic-trajectory flights were also performed. As a result, it was shown that the T-junction is an effective device for roughly separating gas and liquid under microgravity.

View full abstract

Performance of a two-dimensional diffuser is modified by means of an incoming turbulent wake of a circular cylinder installed upstream of the inlet of the diffuser. The pressure recovery coefficient and velocity distribution are measured under various combinations of the diameter and position of the circular cylinder. Reynolds number based on the inlet height and the velocity at the inlet of the diffuser is 4.7×10⁴. The optimum combinations for improvement of the pressure recovery and reduction of loss are determined. Turbulent separation in the otherwise undisturbed diffuser is suppressed by using the optimum combinations.

View full abstract

Low-solidity circular cascades have been applied successfully as a part of the diffuser system of a transonic centrifugal compressor. Three types of diffusers with and without an inlet passage width contraction or a pinch upstream of a cascade were tested. An appropriately contracted diffuser realized high compressor efficiency at high rotor speed due to the decrement in input power. Effects of a pinch on surge and choke were also discussed. Choke occurred at the throat of cascades of a pinched-type diffuser at lower flow rate than the inducer choke flow in spite of a low-solidity cascade. A pinched diffuser had a stabilizing effect on the diffuser characteristics but only a small amount of gain in the surge margin was obtained owing to the change in impeller characteristics.

View full abstract

Decay process of homogeneous turbulence excited by a turbulence generator was experimentally investigated. The turbulence field initially had apparent anisotropy due to the turbulence generator. It decayed downstream into quasi-isotropic turbulence through two stages characterized by the relationship between turbulence energy q² and invariant II of an anisotropic tensor. The return to isotropy proceeded in the first stage and the viscous dissipation dominated in the second stage where the turbulence still maintained weak anisotropy similar to conventional grid turbulence. A new method of determining the virtual origin was proposed and a power law was derived from the equation of turbulence energy employing Rotta's model in order to take the return to isotropy into account. Almost the same decay rate was obtained in both the first and second stages.

View full abstract

Turbulence statistics in the stagnation region of an axisymmetric jet impinging vertically on a flat plate were reported. Particle tracking velocimetry was used for the measurement of highly turbulent flow near the stagnation point. The mean momentum balance was examined to clarify the effect of the turbulent normal stress on the momentum transport in the vicinity of the wall. Distributions of the turbulent stresses and the triple correlations of the velocity fluctuations were presented. The state of turbulence was studied by means of the invariant map of the turbulent stress anisotropy. The budget equation for the turbulent kinetic energy was evaluated, and it was revealed that the negative production was taking place in the vicinity of the wall. Some other features of the budget were also discussed.

View full abstract

The bursting phenomena of the relaminarizing turbulent boundary layer on a flat plate are experimentally investigated using the VITA technique. First, an attempt is made to obtain a normalized bursting frequency universally applicable to the development process of the turbulent boundary layer, the relaminarization process and the retransition process. The ensemble averages of the streamwise fluctuating velocities are drawn and the characteristics of each process are investigated. No time scales can render the non-dimensional bursting frequencies universally constant. The tendency of the distribution of the bursting frequency in the relaminarization process is different from that in the development and retransition process of the turbulent boundary layer. This presumably supports the previous visualization result, according to which the relaminarization process begins from the outer region. In the ensemble averages of the fluctating velocity, it is considered that the relaminarization changes the ejection and sweep, though it does not particularly attenuate the bursting in the inner layer. This phenomenon is qualitatively explained with the aid of the mean velocity profiles and the mixing length concept.

View full abstract

A numerical analysis has been performed on turbulent flow developing in a circular pipe containing one rod located off-center. In calculation, an algebraic stress model was adopted in order to predict Reynolds stresses precisely, and the boundary-fitted coordinate system was introduced to set the boundary condition. The Reynolds number in this calculation is the same as that in the experiment. Calculated results were compared with the experimental data available. The calculated profiles of secondary flow and Reynolds stresses for developing turbulent flow are shown in detail. As a result of this calculation, it was found that the calculated distributions of the streamwise velocity and friction velocity are predicted well. Moreover, the calculated result suggests that a pair of vortices are generated by anisotropic turbulence in the region between the off-center rod and circular pipe. As a characteristic feature, it is pointed out that the intensity of secondary flow is maximum before becoming a fully developed turbulent flow. The location at which it is maximum corresponds to that of the appearance of a merged boundary layer. In addition to these results, the production terms of secondary flow in the vorticity equation are evaluated using calculated results in order to provide further understanding of the generation of secondary flow.

View full abstract

Direct numerical simulation of a three-dimensional spatially developing turbulent mixing layer with mixing transition was conducted to investigate the structures of the mixing layer and the budgets of the transport equations of kinetic turbulent energy and Reynolds stresses. In the transport equation of (2, 2) component of Reynolds stress, the pressure diffusion term has a large value compared with the other terms, whereas it is very small in the budgets of (1, 1) and (3, 3) components. These results suggest that the model for the pressure diffusion term is very important for predicting turbulent flows with high accuracy. The influence of structures in the turbulent mixing layer on the budgets was also discussed. The production of kinetic turbulent energy and Reynolds stresses is governed by large-scale structures and streamwise vortices even after the transition. However, the dissipation of kinetic turbulent energy after the transition is dominated by small-scale vortex tubes in large-scale structures and streamwise vortices in the braid regions. The dominant structure for the transport mechanism and the redistribution mechanism of turbulent energy change from large-scale structures to small-scale vortices with mixing transition.

View full abstract

The vorticity-vector potential formulation method is combined with the 4th-order accurate centered finite difference method without numerical dissipation. The flow in a square cavity is solved at large Reynolds numbers in order to verify effectivity of the computational method. This method is shown to be useful for solving the flow accurately with a small number of grid points as compared with the 2nd-order accurate method. The present method also resolves well the unsteady motion in a square cavity at the Reynolds number of l0⁴. The dynamic subgrid scale turbulence model is implemented in the present computational method. The LES of transition to turbulence in a plane channel is carried out. It is shown that vortices initially roll up from the Tollmien-Schlichting waves, spike and tear, and finally form an array after complex vortical structure. The turbulence statistics are compared with the DNS and the experimental values, and considerable agreement is obtained.

View full abstract

An initial distortion of anomalous extrudate of polyethylene when wall shear rate increases has been categorized as "sharkskin". We examine two distinct extrudate distortions in this study. One is locality of wrinkles in the circumferential direction of the extrudate, which occurs when the temperature of polyethylene is increased from that before the experiment. The other is generation of plugs overlapped with the wrinkles, which occurs when the temperature is decreased from that before the experiment. They suggest a memory effect which appears frequently in a flow of viscoelastic fluid. Two kinds of shear stress-shear rate characteristics appeared and they merged into one line when the wall shear rate was increased and the well-known sharkskin phenomenon was then generated. Frequencies of the wrinkles and the plugs were measured and some qualitative discussions were given for theoretical treatments which will be performed in the future.

View full abstract

Electromagnetic control of characteristics of a low-pressure plasma jet impinging on a flat plate in a pipe is experimentally investigated. The electron density and the electron temperature are measured by Langmuir probe. The heat flux is measured by a cylindrical heat flux probe. The response of plasma current to the bias electric field on a flat plate is measured by an electric probe similar to the Langmuir probe type. An impinging plasma jet near the flat plate area is visualized by means of CCD camera and a thermo-image processor. The distributions of the relative temperature on the flat plate are visualized by means of the interference fringes of chrome oxide on the stainless steel. It is shown that the electron density decreases close to the flat plate in both cases with and without a magnetic field. The region of Maximum electron temperature approaches the flat plate upon application of a magnetic field. The excited atomic temperature distributions show the same tendency as the electron temperature distributions. These characteristics occur due to the active recombination between ions and electrons near the plate and further wall recombination. The response of plasma current is more sensitive to the positive bias under the magnetic field. The relative temperature distributions clearly show that the plasma jet is stabilized by a magnetic field. These findings imply that an impinging plasma jet can be controlled effectively by applying magnetic field.

View full abstract

One of the authors has shown a new hydraulic oscillation. A three-dimensional turbulent jet oscillates periodically when it is issued into an abruptly expanded flow passage, and this oscillation is also caused in the case of gas-liquid two-phase flow. Using this phenomenon, the flow rates of not only single-phase flow but also gas-liquid two-phase flow can be measured. This oscillatory phenomenon is caused in an abruptly expanded flow passage with end plates, but by issuing a three-dimensional liquid (water) jet into an abruptly expanded flow passage without an upper end plate, i. e., an open channel, the jet also oscillates periodically. The condition under which the oscillation is caused, the relationships among the oscillatory frequency, configuration of flow passage, depth of water, flow rate, and others will be clarified.

View full abstract

New self-induced free-surface oscillations were discovered, which were caused by the interaction between a jet, a free surface and a structure. An upward round jet was injected into a cylindrical tank filed with water. The jet impinged on a cylindrical rod (UIS) which was set just beneath the free surface of the tank. Under a certain condition of jet velocity and UIS depth, two self-induced free surface oscillations with their natural frequency were observed. One had no diametric node and two nodal circles, (0, 2), and the other had one node in each direction, (1, 1). (0, 2) mode sloshing was classified into two modes according to their flow patterns. In the (0, 2) mode with outward surface velocity, (0, 2)_OUT the oscillating frequency depended on the jet velocity. In the (0, 2) mode with inward surface velocity, (0, 2)_IN, the mechanism was thought to be similar to that of self-induced sloshing in the previous study. (1, 1) mode oscillation with swelling of the free surface, (1, 1)_α, is found to be the same as "Jet Flutter" in a cylindrical tank without UIS.

View full abstract

A sloshing can be induced by a jet without any external oscillatory force. In a rectangular tank having a vertical plane jet, the first-mode sloshing was observed under a certain condition of the inlet jet velocity and the water depth, i. e., a sloshing condition. In this study, a growth mechanism of the self-induced sloshing caused by the vertical plane jet was theoretically evaluated. A new growth model was proposed, which was based on a feedback loop between the sloshing and the jet. When an oscillation energy supplied for the sloshing was positive, the sloshing was able to grow. The calculated sloshing condition agreed qualitatively with the experimental sloshing condition for different tank geometries. Therefore, the validity of the sloshing growth model was confirmed. The sloshing was found to be caused by the feedback loop constituted by the jet and the free surface.

View full abstract

The breakup mechanism of an annular liquid sheet jet in a cocurrent gas stream is experimentally investigated for gas velocity U_g 30 m/s and liquid velocity U_l 4.0 m/s. A flow visualization technique using a digital high-speed video camera and digital image processing is developed. The annular liquid sheet in the cocurrent gas stream oscillates periodically and is elongated in the downstream direction, and sometimes perforations on the annular liquid sheet surface are observed. Ligaments and rims are formed after the spread of the perforation as the typical configurations of liquid sheet deformation. In the final stage, the liquid ligament breaks up into fine liquid drops as the result of unsteady wave growth on the surface owing to the Kelvin-Helmholtz instability. The basic process of the annular liquid breakup in the cocurrent gas stream is clarified experimentally.

View full abstract

A distributed parameter model of a collapsible tube, in which no effect of separation occurring in the collapsed part is considered, is numerically analyzed by means of the SIMPLE method. Transient flow behaviors, stability of steady flows and self-excited vibrations are investigated. Furthermore, the wave propagation in the tube is indicated as a 3-dimensional bird's-eye view with the temporal and spatial axes. The numerical results are consistent with experimental results.

View full abstract

In front of a reciprocating piston in a pipe containing fluid, the velocity profile is assumed to vary continuously in the axial direction from a uniform one to that of an ordinary oscillatory viscous pipe flow. The present paper deals with theoretical and experimental discussions of this subject. Numerical calculations have indicated that the transition of velocity profiles is completed within a pipe radius away from the piston. In order to experimentally verify this finding, an LDV with a counter-type signal processor is used to measure velocities in oscillatory oil flows close to a reciprocating piston. Although a counter is usually not used for measuring unsteady flows, it has proved applicable to periodic flows when the data within the same range of phase angle are separately collected and stochastically processed in the same way as in an ordinary steady measurement. The measurements thus carried out have given satisfactory results which agree with the theoretical predictions.

View full abstract

It was determined that the vibration amplitude of the tail car is greater, especially in a tunnel, than that of the other cars in a high-speed train, on which aerodynamic force exerts various effects. However, few aerodynamical studies have been conducted and little knowledge has been obtained. To clarify the aerodynamical effects on the train, we carried out a three-dimensional unsteady Navier-Stokes simulation. In order to verify the simulation results, a wind-tunnel experiment was performed. The pressure distributions were compared between numerical and experimental results and running test data available. Surface flow patterns were also compared between numerical and expermental results. These showed good agreement, verifying the simulation results.

View full abstract

Direct numerical simulations of Taylor-Couette flows are conducted to clarify the axial wave number selection of two modes which have different growth rates in the Taylor-Couette flows. By investigating the energy transfer functions, the following conclusions are obtained. (1) In the initial stage, two modes which have different wave numbers grow independently, and the growth rates are determined based on the linear stability theory. However, the mode which is finally selected cannot be predicted using the linear stability theory. (2) The wave number selection of two modes, k₁ and k₂, depends on the nonlinear interaction with the wave number (k₁-k₂). The energy transfers to (k₁-k₂) from k₁ and k₂ play important roles in this process. The ratio between the energy transfer to (k₁-k₂) from k₁ and from k₂ is k₂ : k₁. (3) If the sum of energy transfer from mean flow, viscous dissipation and energy transfer due to the nonlinear interaction becomes negative for one mode, this mode is not selected in the final stage. The energy transfer due to the nonlinear interaction is determind by the energy level at the time when the nonlinearity becomes important.

View full abstract

We took into consideration the classification efficiency in our experiments of an almost rigidly rotating flow between the rotating housing with a spherical wall and the rotor with combined spherical and cylindrical walls. The conclusions are as follows. (1) We obtained the highest sharpness index value of 1.25, which is comparable to that of the recent centrifugal classifiers, in spite of the low rotation number. (2) The values of the cut size and the sharpness index decrease with decreasing Ekman number (rotation number) or Rossby number (through-flow rate). (3) The dispersion rate of feed particles is improved by installing a dispersion nozzle and an orifice near the classification field.

View full abstract

A numerical analysis on the classification of fine particles in an almost rigidly rotating flow is reported. The almost rigidly rotating flow between a coaxially rotating housing and a rotor can be regarded as a system for classification. The flow consists of an Ekman layer, an interior region, a Stewartson layer and a rigidly rotating region. The numerical method is composed of two stages. The first stage involves numerical simulation of the flow field using Navier-Stokes and continuity equations. The second stage involves computation of fine particle motion through the almost rigidly rotating flow. The mechanism of classification of the system is considered using the numerical results. The partial fractional efficiency of the system is computed for various particle diameters and flow parameters ; then the cut size and sharpness index of the system are estimated.

View full abstract

Finite difference solutions to the Doi equation without closure approximations in shear flow are obtained. The Maier-Saupe potential is used to model the mean-field potential which induces the nematic state. Under equilibrium conditions, an isotropic phase is stable when the nematic potential intensity U is less than 4.49 and the lowest value of U for which a nematic phase is stable is 5. In the presence of shear flow, three different types of rheological behavior exist depending on the magnitude of shear rate: aligning, wagging and tumbling. With increasing shear rate, in the aligning regime, the preferred angle has a maximum and the order parameter shows a monotonic increase. For U=6, the transitions from tumbling to wagging and wagging to aligning occur at shear rates of approximately 16 and 32, respectively. The first normal stress difference has a negative value in moderate shear rates which include the whole regime of wagging and a part of the tumbling and aligning regimes.

View full abstract

This paper describes a method by which axisymmetric nozzle flow can be analyzed quantitatively from Mach-Zehnder interferograms. The method applies image processing to obtain fringe shifts from the interferograms rapidly. Using a Laplace transform, the density distributions are calculated from the fringe shifts. In order to calculate the density in the shock wave region, a step function is used to describe the fringe shifts. Pressure can be calculated from the density assuming an isentropic change. The results are compared with those from measurement and numerical simulations, and good agreement is found.

View full abstract

This paper describes damage to living tissue cells induced by plane shock waves. It is an effective method for sensing injury to living tissue cells using pseudo-living cells. In this study, micro-capsules are used to estimate the degree of damage when subjected to shock waves. Mathematical models that can analyze displacement of living tissue cells and pressure in fluid when subjected to shock waves are evaluated. An elastic spherical shell is modeled to fit the biological data. Using stationary and transient analysis of spherical shell, the dynamic characteristics of the living cells (red blood cells and the micro-capsules) are obtained. The results show that damage to living tissue cells depends on (1) the elastic modulus of the cell, (2) the bulk modulus of intracellular material and (3) the thickness of cell membrane.

View full abstract

Recently, there have been some studies on reduction of fluctuating fluid forces acting on a bluffbody by controlling a small portion of the separated shear layer from the body. However, the characteristics of the separated shear layer in the vortex-generating region, which are key information to determine an effective way to control the fluid forces, have not been clarified yet. In this study, downstream behavior of small vortices generated in the separated shear layer near the separation point is investigated by computer image processing of flow visualization. In particular, coalescence of the small vortices at the phase when a new vortex is starting to generate and at the phase when the growth of the vortex is in the intermediate stage are investigated in detail. Also the velocity field mapped by LDV measurement is compared with the visualized image at the same phase to clarify the mechanism of the generation of new vortices.

View full abstract

This paper describes the O₂ LIPF (laser-induced predissociative fluorescence) method for visualization of high-speed flows. Under the condition of weak laser excitation, the linear dependence of O₂-LIPF both on number density of oxygen molecules and on laser intensity is presented theoretically and verified experimentally using a static condition. The vibrational and rotational temperature distributions along a centerline of a supersonic free jet are calculated theoretically, leading to the theoretical fluorescence signal distribution. To improve the S/N ratio of a detection system including an image intensifier and a high-sensitivity CCD camera, three spectral filters are selected to eliminate the scattered laser beam, and three images visualized through them are compared. A low-speed oxygen flow is also successfully visualized using O₂-LIPF.

View full abstract

The effect of the structure and behavior of vortex rings on jet noise containing discrete frequency has been studied by digital image processing. In particular, the behavior of a series of vortex rings released from a nozzle with divergent exit angle was clarified by a high-speed video camera including Schlieren systems. The structures of vortex rings were classified into the following phenomena : chasing phenomenon in which a leading vortex ring was chased by a released vortex ring, merging phenomenon in which a released vortex ring merged after passing-through to a leading vortex ring, and rolling up phenomenon in which the main flow was rapidly rolled up after contraction. It was found that the discrete frequency was affected considerably by the vortex structure and behavior, and the characteristic frequency distributions for the above phenomena were presented. Furthermore, in spite of considerably lower pressure, violent discrete frequency and screech noise such as that generated by a supersonic jet were generated.

View full abstract

Values of overall relative scalable loss δ and of loss distribution coefficient V are analytically determined by extracting from the model performance diagrams of high-specific-speed bulb turbines under a wide operating range. Prior to extracting the scalable loss, the values of guide vane shock loss and of draft tube loss are determined from the model performances at the reference operating points where the nonscalable loss is minimum. Since the shock and swirl losses as well as the frictional loss of runner have been reported in preceding papers, the variation of δ and V can be illustrated for a wide range of setting angles of runner and guide vanes.

View full abstract

Flow pattern and performace of a return channel model were investigated experimentally, as the preliminary step to optimize its profile. The flow angle suddenly changes close to the casing wall near the return channel inlet, and the low energy fluid in this region is carried to the suction side of the return vane. Its fluid finally accumulates over a wide region in the hub on the suction side. This process gives rise to marked the hydraulic loss. Moreover, two kinds of flow pattern appear in the channel outlet section. These flow patterns are maintained in the U-turn bend and meet the stay ring with negative incidence angle. Therefore, the shock and mixing losses cannot be disregarded in the stay ring. The flow pattern, however, becomes uniform at the next-stage runner inlet.

View full abstract

Instantaneous pressures were measured on the casing wall of a guide vane passage and volute casing of a diffuser pump with three different radial gaps between the impeller exit and guide vane inlet. The measured pressure fluctuations were found to decrease with increasing radial gap. Using our method, numerical calculations were performed for the pressure fluctuations, and the calculated pressure fluctuations were found to agree with measured ones. Moreover, the ensemble-averaged pressure data show the effects of the radial gaps on the spatial distribution of unsteady pressures, which causes the rotation of the circumferential pressure mode due to rotor-stator interaction.

View full abstract

A theoretical study was carried out on transient behavior of a cavitating pump system during sudden opening or closure of the discharge valve. Assuming that the transient pump performance is quasi-steady, the system dynamics were determined in a simple fashion for the quick valve opening case. A characteristics method was applied to the transient behavior in the quick valve closure case under the assumption of quasi-steady pump performance. Calculated pressure as well as flow rate at transient agreed well with the measured values. Moreover it is shown that unsteady fluctuations of pressure and flow rate are caused by cavitation behavior during a sharp increase in flow rate, and mainly by a water hammer with water column separation during a sharp decrease in flow rate.

View full abstract

This study concerns a new application of the Bond graph to the analysis of dynamic behavior in a turbomachinery system. A theoretical and experimental study has been made on the transient characteristics of a centrifugal pump during the starting/stopping periods. Instantaneous rotational speed, torque, flow rate and total pressure rise are measured in the experiment. Theoretical calculations for the prediction of transient characteristics using the Bond graph, which is a one-dimensional analysis of the system dynamics based on energy conservation, are developed and the results are compared with the corresponding experimental results. It was clarified that the results of Bond graph simulation agree well with the experimental results, and that Bond graph simulation is very useful for predicting the pump transient characteristics during starting/stopping periods.

View full abstract

An experimental study was made to clarify the fundamental flow characteristics of Scirocco fans. It was found that the impeller generally suffers from instabilities caused by the positive slope of the Euler head when it is used without a casing. This instability can be distinguished from that of impeller rotating stall. The instability can be suppressed by using a casing. Thus, another important role of a casing, other than pressure recovery, has been elucidated.

View full abstract

Bubbles generated in conductive ink heated directly by electric current pulses can be used in an ink-jet printer to eject ink-droplets. Such parameters as the time required for bubble nucleation, growth rate of the bubble, maximum radius of the bubble and electric power input required for bubble generation, which depend strongly on the shape and size of electrodes immersed in the ink, influence the performance of the printer. A method of numerical analysis was developed to investigate the temperature distribution of the ink and dynamics of the bubble, and was confirmed to give values of the above parameters which were in good agreement with those of experiment. An optimum configuration of the electrodes was suggested, based on the comparison of results calculated for different shapes and dimensions of the electrodes.

View full abstract

The general expression for heat of vaporization of a mixture at constant pressure, that at constant temperature, and that at constant pressure, temperature and composition are derived. The third expression is related to the liquid-vapor interface where steady vaporization or condensation is taking place. Numerical examples are shown for binary mixture of HCFC(R)22-HCFC(R)123.

View full abstract

The evaporation process of a suspended droplet was examined for several kinds of paraffin hydrocarbon fuel, distilled water and ethanol in a hot turbulent air flow. As the turbulence intensity increases, the evaporation rate constant increases for all materials because the Kolmogorov microscale becomes smaller. When the boundary layer thickness of vapor concentration around a droplet is thin, the droplet evaporation is accelerated as the mass transfer rate of vapor is increased by the flow turbulence due to steep gradient of the concentration. The vaporization Damkehler number is not suitable for estimating the influence of turbulence on evaporation rate under the condition where evaporation occurs intensely.

View full abstract

Experiments were performed to investigate the flow patterns which occur in upward air-water two-phase flow in vertical twisted-tape-inserted and circular empty tubes. Three basic flow patterns, i. e., bubbly, intermittent and annular flow, were defined based on the observations in the present experiments. At high water flow rate and small twist ratio, the bubbly-intermittent transition occurred at a lower air flow rate in the twisted-tape-inserted tubes than in the empty tube due to the centrifugal force of swirl flow which concentrated bubbles at the center of tube. The effect of centrifugal force was successfully modeled to obtain the criterion equation of the bubbly-intermittent transition. The intermittent-annular transition could be predicted well by the existing criterion equation developed for empty tube.

View full abstract

A quantitative analysis of critical heat flux (CHF) under high mass flux with high subcooling was successfully carried out by applying a new transition region model for a thin liquid sublayer on the heated walls to the existing model of vapor blanket over the thin liquid sublayer. The CHF correlation proposed in this study could predict well experimental data obtained using circular tubes of 2.5 to 3 mm in diameter, 100 to 290 mm in length, 0.34 to 2.6 MPa and 13 to 210°C in inlet subcooling, and rectangular channels of 3.5 mm in gap, 150 mm in length, 0.5 to 3.1 MPa and 13 to 166°C in inlet subcooling, revealing a unique feature that the critical heat flux is proportional to around G⁰.8, where G is the mass flux, at high mass flux of the dimensionless mass flux G^* ranging from 350 to 4700 under the subcooled condition at the exit of the channel.

View full abstract

Critical heat flux in an open two-phase thermosyphon has been measured employing water, R113 and R22 for a wide range of L/D=4.8 to 960 and density ratio of ρ_L/ρ_G=6.17 to 1602. The CHF data measured are found to be in satisfactory agreement with those calculated from the maximum liquid falling rate in countercurrent annular flow of liquid and vapor. Normal operation of the thermosyphon cannot be obtained for a small tube diameter of less than 2 mm.

View full abstract

In this report, subcooled film-boiling heat transfer of liquid film flow formed by liquid spray on a hot surface was simulated by film-boiling heat transfer of a subcooled liquid film flow supplied from a nozzle along the hot surface. Experimental data indicated that the effects of film thickness, wall superheat and distance from the leading edge on the film-boiling heat transfer were weak. The main parameters controlling the film-boiling heat transfer were liquid velocity and subcooling. The modified heat transfer coefficient was defined as the ratio of wall heat flux to liquid subcooling and a heat transfer correlation between the Nusselt number based on the modified heat transfer coefficient and the Reynolds number was obtained. The correlating equation was very similar to that of turbulent heat transfer of single phase flow, and it was found that most of the heat flux was dissipated to increase the liquid temperature.

View full abstract

Forced-natural-mixed convection heat transfer characteristics of a one-stage spherical particle layer in an air flow are investigated experimentally. The one-stage spherical particle layer is provided in a rectangular cavity having a heating bottom surface and is installed at the lower part of a rectangular air channel. Three types of spherical particles, which have almost the same diameter (about 10 mm) but different thermal conductivities, are tested. The cavity length is varied from 19.6 mm to 90 mm, and the cavity depth is varied between 0 mm and 10 mm. The variation of the heat transfer coefficient with the length of the heating surface can be explained by the flow behavior around the particle layer ; that is, the case of the zero cavity depth, the flow is suppressed by the particles within about three rows from the leading edge. Therefore, the heat transfer coefficient increases with a derease in the length of the heating surface. In the case with a shorter heating surface (about twice the diameter of the particle), a considerable increase in the heat transfer coefficient is achieved because the flow from the upstream runs through the particle layer. In the case of the cavity depth having almost the same value as the diameter of the particles, air flows over the particle layer and air near the cavity bottom is almost still. In this case, a decrease of the length of the heat transfer surface decreases the extent of the flow in the depth direction, and the heat transfer coefficient is also decreased. The heat transfer coefficient can be expressed as a nondimensional heat transfer correlation, where the effects of the cavity length, the cavity depth, the air velocity and the temperature difference between the air and the bottom surface of the cavity are taken into consideration.

View full abstract

An analytical model is developed to calculate the radiative heat transfer through a fibrous layer with various fiber orientations consdering the effect of the diameter and the optical properties of the fiber. Milne-Eddington approximation is adopted to obtain a two-flux model type of formulation. To facilitate determination of the extinction efficiency, scattering albedo of the fibrous layer and backscatter factor required for the analysis, these values are shown graphically as the functions of complex index of refraction, size parameter and fiber orientation. The validity of the present model is shown by comparing the results with ones obtained by the Monte Carlo method.

View full abstract

The Boussinesq approximation has been commonly used to analyze thermally driven flows in enclosures. However, its validity is limited to the case of small temperature difference. As temperature difference increases, the practical flow characteristics and instabilities show increasing differences from those obtained with the Boussinesq approximation. In the present study, we have proposed modified governing equations of thermal convection which can well represent the flow behavior in enclosures with relatively large temperature difference. Also, we have discussed How the flow pattern in a square cavity varies with increasing temperature difference and How the flow instability occurs, comparing the numerical results obtained using the Boussinesq-approximated and the exact governing equations.

View full abstract

The Marangoni convection in a supported liquid droplet suddenly subjected to radiative heating by a hot environment is analyzed by two-dimensional numerical analysis. The effects of the thermal conductivity and the heat capacity of the support rod on the flow and temperature rise within the droplet were studied. The following results are obtained. (1) The surface-tension-driven convection occurs because of the difference between the temperature rise of the surfacees of the support rod and the liquid droplet. (2) The Marangoni convection within the liquid droplet decreases when the heat capacity or the thermal conductivity of the support rod becomes small. (3) The mean temperature change within the droplet is scarcely affected by the existence of the support rod, but when the heat capacity or the thermal conductivity of the support rod is large, the mean temperature becomes somewhat smaller than that in the case with out the support rod. (4) The maximum temperature of the droplet becomes small when the heat capacity or the thermal conductivity of the support rod is large.

View full abstract

By adopting results from past research concerning the freezing phenomenon of supercooled water and the method to control the initiation of the freezing, a method for measuring thermal conductivity of water below the freezing temperature was introduced. The method was basically a non-steady-state hot-wire method, and a few modifications were made to avoid freezing during the measurement. It was found that the thermal conductivity of supercooled water can be measured by considering factors such as selection of vessel material, size and shape, reduction of the complexity of the whole instrument, and purity of water. Consequently, thermal conductivity was obtained at 265K. According to the results obtained, the thermal conductivity decreases gradually with decreasing temperature, and follows a tendency similar to the temperature dependence at the region higher than the freezing point.

View full abstract