A state-of-the art review is presented on the research activities in Japan on condensation heat transfer during these nine years. On writing this article the author has decided to choose those papers which have been published on the Trans. JSME from 1980 until 1988 as the main target of the review. The papers printed elsewhere on the other publications or prior to 1980 are referred to only whenever it is needed. The author classifies the subjects into five items. They are ; (1) Film condensation of single-component vapor, (2) Film condensation of multi-component vapor, (3) Enhancement of condensation heat transfer, (4) Dropwise condensation, and (5) Direct contact condensation and other forms of condensation. However, the author's effort is focussed mostly on the item (3) on the techniques of enhancement of condensation heat transfer. Dropwise condensation is another item that is discussed in some detail. Only the outlines are presented for the remaining items. Also it should be noted that the papers referred to in this article do not cover all the publications on condensation heat transfer even for those published after 1980 on the Trans. JSME, though more than 100 references are cited.

View full abstract

Water tests in a cylindrical vessel have been performed to investigate the turbulent transport phenomena in the mixing shear layer which moves upwardly and affects the structural integrity of the vessel and internal structures during a transient condition of a fast breeder reactor. The turbulent characteristics of the coolant adjacent to a thermally stable stratified interface have been clarified by using the technique combined with 3 beam LDV and thermocouple probe. The correlations among two-component volocities and temperature are obtained simultaneously. From the experimental results, it was found that the maximum temperature gradient can be realized in the almost laminar-like condition in which the turbulent heat fluxes in the upper and lower layers are both damped.

View full abstract

A developing turbulent flow of power-law fluids in the entrance region of curved pipes has been investigated by measuring axial velocity profiles with an optical-fiber LDV under the conditions of power index n>0.75, curvature ratio of the pipe R_c=10 and 30, and Reynolds number Re≒10 000∼33 000. The turbulent flow grows more mildly and the inlet length becomes longer than the laminar flow. A Reynolds stress model is applied to a fully-developed turbulent flow. Numerical results are in moderate agreement with experimental results within power index n=1∼0.9. In addition, an approximate expression for turbulent friction factor is presented on the basis of the boundary layer theory.

View full abstract

The transition from laminar to turbulent flow in the radial liquid film flow on a rotating disk is studied experimentally. The flow is generated by letting water discharge from a cylindrical opening between a horizontally placed rotating disk and a circular nozzle into the atmosphere along the upper surface of the disk. High speed photographs of the process of transition are presented. Moreover, the influence of the rotational frequency of the disk on the radius occurring the transition is investigated. The results show that, while the rotation of disk suppresses the laminar-turbulent transition as long as the circumferential velocity is relatively small to the radial velocity, it promotes the transition when the circumferential velocity becomes large.

View full abstract

The radial thin liquid film flow on a rotating disk is investigated experimentally. Velocity distributions in the flow are measured by laser Doppler velocimeter and the liquid film thickness is measured by a point gage method. LDV measurement on the laminar flow clarify that a three-dimensional laminar boundary layer grows gradually in the downstream direction and finally the whole flow merges into a three-dimensional laminar boundary layer. Furthermore, it is found that surface velocity and liquid film thickness agree reasonably with Miyasaka's analysis as long as the flow is laminar. It is also found that the experimental results differ considerably from Miyasaka's analysis when the laminar-turbulent transition occurs.

View full abstract

Experiments were conducted on the effect of a tripping-wire on the laminar-turbulent transition of boundary layers over wedges of 15, 30 and 60 degrees with a pressure gradient. The results obtained are as follows : (1) the transition Reynolds number R_t=U_∞x_t/ν at the most effective condition is given as the formula R_t=1.06R_k+1.33×10⁴ for the range of the wedge angle between 15° and 60°, where R_k (=U_∞x_k/ν) is the wire Reynolds number, U_∞ the free-stream velocity, x_k and x_t the location of the wire and the transition point measured from the apex of the wedge, respectively, k the wire diameter, and ν the kinematic viscosity of the fluid. (2) The minimum transition Reynolds number R_tm is a function of the ratio of the height of the roughness element k to the displacement thickness δ* of the boundary layer at the location of the element, and R_tm decreases with increasing k/δ*. R_tm increases with an increasing wedge angle at the same value of k/δ*.

View full abstract

The laminar-turbulent transition of a boundary layer on a rotating flat plate has been investigated. After careful examination of the transition of the stationary boundary layer, the flat plate was rotated around the axis which was parallel to the plate and perpendicular to the free stream direction. The measurement of the wall shear stress and the flow visualization revealed significant effects of Coriolis acceleration on the transition. The transition of the low pressure side boundary layer was characterized by a turbulent spot which had a distinct boundary with the quiescent surroundings. The transition of the high pressure side boundary layer was characterized by enhanced vortical motions.

View full abstract

Contour maps of induced shear stress were three-dimensionally demonstrated about a large-scale horseshoe vortex artificially induced in a fully developed turbulent boundary layer. Two kinds of shear stress components were induced in the horseshoe vortex : the shear stress inherent to the coherent fluid motion and that generated through the interaction between the coherent motion and background turbulence. More than 70% of the induced shear stress was concentrated about the symmetry plane of the horseshoe vortex corresponding to the strong ejection, where the coherent component acted a dominant role. In the outer side of the legs of the vortex, the sweep motion induced the negative shear stress, to which the interaction component contributed most.

View full abstract

The flow in a two-dimensional channel with repeated square ribs was investigated numerically. A stream function-vorticity formulation for the flow equations was used, and was solved by the finite difference method using the pseudo-unsteady technique. The calculation was performed for the range of Reynolds number of 5≤Re≤3 000, and for the range of pitch ratio of 1.8≤W/h≤11. The results showed that the flow structure changed with both variations of the pitch ratio and the Reynolds number. In this paper, particular attension was paid to the flow pattern over the roughness elements, the influence of roughness on the flow over the roughness and the behavior of vortex in a groove of roughness element. The flow pattern obtained from the calculation was classified into three cases which were named the one vortex type, the semi-two vortex type and the two vortex type, respectively.

View full abstract

The objective of the study is to examine numerically the relation between the flow structure and the dynamical characteristics for the laminar flow in a two-dimensional rough channel. As a rough channel, the two-dimensional channel with repeated square ribs is considered. The Navier-Stokes equation and the equation of continuity have been solved numerically by the finite difference method. The calculation was performed for the range of pitch ratio 1.8≤W/h≤11, and for the range of Reynolds number 5≤Re≤3 000. The results showed that both the shear stress and the pressure acting on the roughness elements changed with the variation of flow pattern over the roughness elements depending on the pitch ratio W/h and Reynolds number Re. The contours of the ratio of pressure drag to total drag acting on the rough wall was shown in the manner of the division chart in W/h Re plane.

View full abstract

Fully-developed turbulent flow in a channel rotating about a spanwise axis is calculated using a Reynolds stress model. The model is a slightly modified version of the second-moment closure applicable right up to a wall which was developed by the author in previous papers. The numerical solutions are compared with available experimental data on the mean velocity profiles and the wall shear stresses and also with large eddy simulation (LES) results on the turbulent normal stresses. When the Rossby number is lower than about 0.1, the predictions are in good agreement with the experimental data and the LES results, capturing the stabilizing and destabilizing effects of Coriolis forces on the flow. For Rossby numbers greater than the value, however, the calculations fail to reproduce the experiments, which suggests a need for a separate study.

View full abstract

The turbulent flow mechanism in a wall-attachment fluidic device was analyzed with numerical predictions. The computation was carried out with the finite difference method in conjunction with the k-ε turbulent model and the semi-implicit method for the pressure-linked equation. The computation procedure employed in this paper could define the boundary conditions definitively at the ports for input and output. The numerical solutions provide the velocity profiles, the pressure distributions, the turbulent energy distributions, and the energy dissipation rate distributions in the flow field. The switching flow rate obtained by the numerical computation was compared with that measured by experiment. It is found that the numerical predictions closely simulated the flow mechanism during the process of the jet switching in the device.

View full abstract

Turbulence measurements are performed in curved circular tubes for three values of the ratio of coil to tube radius, R/a. Distributions of turbulence energy and axial and radial velocities are obtained in the fully developed flow with a hot wire anemometer. Turbulence energy and Reynolds stress become maximum at the position in the core region where turbulence generation becomes maximum. Turbulence generation and dissipation does not balance because of the convective effect induced by the secondary flow when R/a is mrall. Experimental results of the ratio of Reynolds stress to turbulence energy are compared with those of theoretical analysis.

View full abstract

An axisymmetric turbulence field was generated behind honeycomb grids in a wind-tunnel test section. Integral and microscales were experimentally determined and their streamwise evolution was inspected. The aspect of the evolution differed from that of the quasi-isotropic turbulence, because it was influenced by the energy exchange between the spectral components of velocity fluctuations. The turbulence Reynolds numbers based on the integral scales became regardless of the degree of the anisotropy if the length scale was taken as the geometrical mean of the longitudinal and lateral integral scales. The turbulence Reynolds number based on the equivalent microscale was not free from the effect of the streamwise deformation of the spectra. These facts issue a warning to the easy application of the 'local-isotropy concept' on the anisotropic turbulence.

View full abstract

Using the regular grid system, an algorithm for the numerical solution of a Navier-stokes equation accompanied with the presently developed schemes is proposed. After being developed into body-fitted coordinates, NSMAC is applied for the analysis of laminar flow in a duct with divergence and convergence to confirm that NSMAC is free from checkerboard pressure and residual error of mass continuity. Further, NSMAC is applied for the analysis of turbulent flow between two parallel plates by using the k-ε model, where a good agreement between calculation and observation and its high CPU time performance are confirmed.

View full abstract

When a viscoelastic fluid flows out from a horizontal duct, many cracks and protruding ridges are formed on the jet surface. In this paper, the effects of gravity and fall length of the jet on the anomalous phenomenon are investigated. Firstly, we shown that the anomalous phenomenon does not appear on jets extruded into water. Secondly, we clarify that the appearance critical shear rate of the anomalous phenomenon increases with an increasing angle of inclination of the duct which is measured downward from the horizontal plane and with a decreasing fall length of the jet. Moreover, making use of the dimensional analysis method considering the effects on these various factors, we find an empirical formula for the appearance critical point of the anomalous phenomenon.

View full abstract

There has been much research on the physical properties of magnetic fluids. However, investigation of constitutive equations has been far from adequate. In this paper, a new and complete set of fundamental equations for magnetic fluids having good conductivity of heat and electricity is formulated, considering viscoelastic effect by thermodynamical methods. The quasi-conservative part and dissipative part of constitutive equations are determined by taking partial derivatives of free energy and dissipation function, respectively. New fundamental equations are more general and strict in comparison with previous ones. Introducing a concept of electromagnetic exergy, this paper separetes pressure into two parts, on from electromagnetic effects and the other from thermodynamical effects. It also demonstrates how electric and magnetic effects act independently each other to increase apparent viscosity, and proposes new dimensionless numbers which represent these effects.

View full abstract

In this paper, the influences of various factors on the drop formation and the behavior of drops are experimentally examined in a uniform drop region in which ink drops are released very regularly form a capillary nozzle. For that purpose, the outside diameter of the nozzle (the inside diameter is constant), the inside diameter of the nozzle (the outside diameter is constant), the magnetic force per unit volume and flow rate are widely varied. In addition, liquids having different values of surface tension are used. Then, various factors governing the period of drop formation and the drop diameter are found by using dimensional analysis. The empirical formula for the dimensionless period of drop formation and the dimensionless drop diameter are given.

View full abstract

This paper describes the steady two-phase flow field of supersonic ges-particle freejets issuing into quiescent gas, which has been numerically analyzed. The computations use a time-dependent technique using the TVD-MacCormack algorithm for both gas and particle phases. The freejets considered here are axisymmetric and inviscid, and single-sized spherical particles are considered. The study focuses on the influence of not only loading ratio, but also particle size upon the freejet structure. Computational results show that the particles affect the freejet structure significantly. Specifically, the position of a Mach disc moves downstream when the particle size decreases and/or the loading ratio increases.

View full abstract

Three-dimensional structures of two- interacting free jets with different source pressure are studied by flow visualization. Angles between jet axes are set at θ=45° and 180°. Experiments are carried out for various source pressure ratios and various ratios of the source pressure to the pressure in the expansion chamber. For θ=45°, many types of flow field appear with respect to the source pressure ratio and the structures of these flow fields are classified into several types according to their shock systems. Characteristic three-dimensional structures and a contact surface with bends in an interacting region are explained in detail. For θ=180°, structures of the flow field are divided roughly into four types depending on the condition of the source pressure and the pressure in the expansion chamber. Specifically, stable and unstable flows in the interacting region are found for the equal pressure ratios.

View full abstract

Numerical simulation of the steady two-phase flow field of gas-particle inside a JPL nozzle is described in this paper. The computations use a time-dependent technique using the TVD-MacCormack algorithm for both gas and particle phases. The nozzle flows considered here are inviscid, and spherical particles of 1-5μm in diameter are treated. Numerical results show that the nozzle expansion is delayed as the interaction between gas and particle phases becomes strong. In addition, it is seen that there exists a particle-free zone in the divergent section of the nozzle.

View full abstract

An experimental study has been conducted on a fully developed turbulent flow through a rectangular duct with one roughened short-side wall. Detailed results have been obtained for mean velocities and turbulent stresses using hot-wire anemometers. One large secondary-flow vortex appeared near the corner formed by the smooth and roughened walls ; it was similar to that obtained in the rectangular duct having one roughened longer-side wall. Near the corner formed by the two smooth walls, the secondary-currents proceeded for the corner like those ion a smooth walled duct. Furthermore, it was found that in the rough duct, the contribution of turbulent shear stress (u₂u₃)^^^- was as important as that of anisotropy of cross-planar normal stresses ((u²₂)^^^--(u²₃)^^^-) for the generation of the secondary-flow of the second kind.

View full abstract

A bouncing model and a particle source term model developed by the authors for two-dimensional channel flows were extended to three-dimensional flows. A numerical simulation for a horizontal pipe flow was conducted with the present models to predict the properties of flow (particle concentration, velocities of particle and gas, pressure drop). The numerical results were presented for the case where polystyrene particles of diameter 0.41∼1.0mm were transported in a plastic pipe of diameter 52 mm. It is found that the numerical results give a good prediction as compared to experimental data.

View full abstract

A gas-solid two-phases flow in a vertical pipe was investigated experimentally. Particle velocity and concentration were measured by an optical fiber probe, and gas velocity by a Pitot tube. Two kinds of polystyrene spherical particles were used, the mean diameters of which were 0.406 and 1.50 mm, and the pipe diameter was 40 mm. It was found that the concentration distribution has a maximum at the pipe center. The distribution of particle velocity is nearly flat for both particles. The distribution of gas velocity is strongly affected by the particles. In order to evaluate the degree of the particle influence on the distribution of gas velocity, a displacement thickness was considered and obtained for the present gas flow. The displacement thickness is found to correlate well with the mean concentration.

View full abstract

Experimental investigations on the characteristics of the fluctuating forces acting on rectangular cylinders with fully separated flow were carried out at a Reynolds number of 5.5×10⁴. The widty-to-height ratio B/H of the section was varied from 0.3 to 2.5. Phase-averaged lift and drag forces were obtained in terms of phase-averaged surface pressure. The phase-averaged fluctuating lift and drag coefficients take the respective maximum values in the same tendency as that of the time-mean drag coefficients for the critical geometry of B/H=0.7. In addition, the torsional moment forces associated with the phenomenon of the vortex-induced flutter were discussed in detail based on the characteristic of the fluctuating pressure, whose frequency equals to the vortex shedding frequency, acting on the surfaces of the rectangular cylinders.

View full abstract

A method of measuring three-dimensional steady flow in an agitated vessel is proposed. The flow field in the vessel is visualized by using tracer particles and a velocity distribution is measured by digital image processing. The three components of the flow velocity are obtained by measuring the velocities in two cross sections perpendicularly intersecting each other. A compound slit with three layers designed by the authors is used to illuminate the measured cross section. A reasonable result is obtained in the measurement of the velocity distribution near an impeller of the agitator. This measuring method is applicable to any other three-dimensional steady flow.

View full abstract

Since the operating range of a Francis turbine is limited by the draft tube surge, the magnitude of pressure fluctuation at the draft tube wall should be examined at the test using a model turbine. However, there is the serious problem that the nature of pressure fluctuation depends on the measuring point which differs point by point even in one cross section of the inlet cone. The present authors attempted to explain it as a result of the difference in phase between two fluctuating components, rotating type and synchronous type, both of which were generated by the orbital motion of a spiral vortex core. Thus, based on this concept, we developed a method to evaluate the two components separately using pressure data measured at two taps which were installed at an angle of 90 deg in the cross section. The experimental results indicate that the proposed method is usable as a diagnosis of pressure fluctuations which are generated by a spiral vortex core in a draft tube.

View full abstract

When the curvature of bend in S-type draft tube is large, the deterioration of draft tube performance is resulted from occurrence of flow separation inside the first bend of S-type draft tube. To overcome this problem, the suppression of flow separation by boundary layer control is discribed in this report. ; (1) Providing a blade near the area where a separation might occure, (2) Providing several small vanes for generating vortices on the wall surface upstream of flow separation.

View full abstract

Boundary layer fences were attached to end walls of a linear turbine rotor cascade, in order to increase the circumferential component of the aerodynamic blade force. Measurements were taken of the static pressure distribution on the channel-confining surface, for different heights and pitchwise locations of the fences. The pressure distribution correlated well with the secondary flow pattern visualized by the surface oil flow method. The fences diminished effectively the crossflow-induced local pressure rise in the tip region of the blade suction surface, but decreased the passage-vortex-induced local pressure drop also on the suction surface. The fences were most effective if they were 1/3 of the inlet boundary layer thickness high from the end wall and located half a pitch away from the blades, reducing the thickness of the force defect at the blade tip by 40 percent.

View full abstract

In this paper, the optimization of single stage axial flow ammonia turbine for OTEC (Ocean Thermal Energy Conversion) is carried out by the Hooke & Jeeves' Pattern Search Method and the Powell Method (the method of steepest descent). The parameters in the objective function (turbine efficiency) consist of specific speed, specific diameter, nozzle outlet angle, ratio of blade to diameter, number of nozzle blades and rotor blades. The results of maximum efficiency of turbine, optimum rotational speed, optimum mass flow rate of working fluid, etc are shown for a turbine of gross power output between 1 MW and 25 MW.

View full abstract

Hydrodynamic force perturbations on a three-blade centrifugal impeller were examined by measuring the fluctuating pressures on the impeller blades and transforming their periodic component into Fourier series. Comparison of the forces between the three-and seven-blade impellers was made and the effects of the blade numbers on the hydrodynamic force were clarified. It is found that there are few differences in the hydrodynamic force exerted on one blade between the two impellers. However, hydrodynamic force on the three-blade impeller obtained by the vector summation of each blade force exhibits a large fluctuation than that on the seven-blade impeller.

View full abstract

To predict the behavior of air-water two-phase flows in a centrifugal pump impeller, a three-dimensional numerical method is proposed based on a bubbly flow model. If it is assumed that the mixtures are homogeneous bubbly flow containing fine bubbles compared with the characteristic length of the impeller channel, than the equations of motion of the mixtures are represented by those of liquid phase and its velocity is expressed as a potential for the quasi-harmonic equation. The equations are solved by use of the finite element method to obtain the velocities and pressures, and the equation of motion of an air bubble is integrated numerically on this flow field to obtain the void fraction. These calculations are repeated until the solutions converge. The results obtained show good agreement with experiments within the range of bubbly flow regime.

View full abstract

This paper is concerned with the determination of the performance change in pumps operating under two-phase flow conditions based on the calculated distributions of velocities and void fraction, in which the mixture is assumed to be an inviscid bubbly flow with the slippage between two phases. The estimated changes in the theoretical head are confirmed with experiments within the range of the bubbly flow regime. The calculated head due to the momentum exchanges between two phases is also favorably compared with the measured value of the head loss within the same range, when the hydraulic loss due to wall friction is assumed to be zero under two-phase flow conditions. The head-loss ratio obtained shows nearly the same value as the experiments, showing this analytical model to be applicable to the prediction of the two-phase flow pump performance.

View full abstract

Based on the evaluation of oxygen transfer performance in terms of non-dimensional variables such as oxygen Sherwood number, fractional saturation and non-dimensional length, the conditions achieving the high performance of gas exchange is discussed. The non-dimensional length at the rated condition, i which the oxygen saturation at the outlet becomes 97%, was determined in the relation between the fractional saturation and non-dimensional length. Simultaneously the oxygen transfer performance expressed by oxygen Sherwood number has been evaluated in relation to the variation of non-dimensional length. Finally the design parameters such as internal diameter, tube length and residence time are determined based on the rated condition.

View full abstract

The experimental study was conducted to investigate the performance of a prototype sprinkler for melting snow or railway tracks at various pressures. The heat loss in flying water droplets and the performance of the sprinkler were theoretically obtained. The mean diameters of water droplets were measured at different water flow rates by photography : they were 5∼6 mmφat the initial speed of 11.5∼16.0 m/s. The temperature drop of the droplet is about 0.2°C/m in flight through air at 0°C. In this sprinkler, the distribution of sprinkled water is kept uniform regardless of the water flow rate.

View full abstract

The wetting behavior of a two-dimensional liquid meniscus fomred under a horizontal plate was investigated theoretically. In the previous report, we considered the wetting mechasism from a macroscopic viewpoint ; that is, we assumed that Young's equation could be applied to the apparent macroscopic wetting problem. In this report, in order to substantiate the assumption made in a previous report, the same problem was analyzed precisely from a microscopic viewpoint when there existed infinitesimal two-dimensional roughnesses on the horizontal plate. The thermodynamic energy change of the system with the meniscus moving along the roughend surface was calculated. When the apparent macroscopic angle between the meniscus and the plate was equal to the advancing or receding contact angle, the system became unstable. This result agreed with that of the previous report. The assumption that Young's equation could be used macroscopically was verified by this precise analysis.

View full abstract

The spectral normal emissivity, reflectance and transmittance of silicon wafers were measured at a temperature of 950°C and wavelengths of 0.7 to 9μm. Samples were seven silicon wafers with oxide films of 2.2 to 628 nm film thickness and nitride films of 50 to 190 nm film thickness with small dopant density. It was found that the emissivity of specular surfaces of silicon wafers with very thin oxide film at wavelengths of 0.7 to 8 μm were 0.6 to 0.7, the reflectances were 0.1 to 0.3 and the transmittance were to 0 to 0.2. The emissivity of silicon wafers with oxide films of more than 75 nm film thickness and nitride films of more than 48 nm film thickness changed from 0.6 to 1.0 according to film thickness and wavelength. The emissivity of the rough surface of a wafer was greater than that of a specular surface by 0.07. The emissivity and reflectance of silicon wafers were calculated using the free carrier absorption theory considering the interference phenomenon. Calculated results agreed with experimental results.

View full abstract

The present paper reports absolute measurements of the thermal conductivity of molten salts by the transient hot-wire method using newly developed ceramic-coated probes. Al₂O₃ has been employed as the insulation material and the following techniques have been utilized to produce insulation layers : ion plating for a thin platinum hot-wire and plasma-spray for thick titanium struts. The measurements have been performed on pure NaNo₃ in the temperature range 311°C to 388°C and KNO₃ 349°C to 439°C. The experimental data have an estimated accuracy of ±3%. In comparison with the present results, most of he previous experimental data obtained by steady-state methods show higher values which may be due to the presence of convection and radiation.

View full abstract

In measuring thermal conductivity of a corrosive or an electrically conductive liquid using the transient line source technique, the wire must be coated. This report analytically investigates the effects of he coating layer on the measured value using the GFD and FDM, considering two cases where the thickness of the layer is uniform in circumferential direction, or unevenly distributed. The calculation is carried out varying the measurement conditions, such as the thermal conductivity of the coating material and of the sample liquid, size of the wire, and thickness of the layer, and the change in the apparent thermal conductivity of the sample liquid with time is presented. It was found that the thermal conductivity of the sample liquid measured is not significantly affected by the coating.

View full abstract

This paper reports measurements of the vapor pressure and PVT properties of the R 115 (CClF₂CF₃). The present measurements were made by the constant volume method coupled with isothermal expansion procedures for a certain fixed density of a sample. The measurements were conducted in the range of temperatures from 296 to 443 K and of pressures from 0.8 to 9.6 MPa and of densities from 235 to 1176 kg/m³ along 11 different isochores. The number of data for the vapor pressure is 68 and that of the PVT data is 94. The experimental uncertainties in the temperature, density and pressure were estimated as ±8 mK, ±0.1 % and ±3 kPa, respectively. We calculated the vapor pressure correlation based on our vapor pressure data and Matthias' vapor pressure data. The vapor pressure correlation represents vapor pressure from 227.5 K to critical temperature. Comparing our vapor pressure data with other reported data and correlations, it was found that the present measurements were consistent and reliable. The present PVT data were also compared with other reported data and available equations of state.

View full abstract

The film condensation of a downward flowing R-113 vapor on in-line bundles of horizontal finned tubes was experimentally investigated. The test sections were 3×15 bundles with or without two rows of inundation tubes at the top. Two tubes with flat-sided fins and four tubes with three-dimensional fins were tested. Heat transfer measurements were carried out on a row-by-row basis. The effect of the vapor velocity on the finned tube bundles was considerable smaller than the case of a smooth tube bundle. The decrease of the heat transfer coefficient due to the condensate inundation was small for tubes with flat-sided fins, while a considerable decrease was observed for tubes with three-dimensional fins. A tube with flat-sided fins of a small fin pitch (0.5 mm) showed the best heat transfer performance over the range of film Reynolds number tested (<2 500).

View full abstract

The local mass transfer has been measured for a circular cylinder placed downstream of and perpendicular to a single row of cylinders in a uniform flow. Using flow visualization techniques and a flow direction probe, and considering the wall pressure distributions, the flow behavior around the cylinder was investigated varying the distance, L, between the upstream and downstream cylinders under the constant Reynolds number, 5×10⁴. The characteristics of the local mall transfer and fluid behavior could be classified into two patterns as follows : (1) At 1.5≤L/D≤2.5, a horse-shoe-like vortex and inner secondary vortex were detected using flow visualization techniques. The local mass transfer distributions show a remarkable three-dimensionality due to these vortices. (2) At 2.5≤L/D≤8.0, the distribution of he Sherwood number were similar to that in the super critical Reynolds number range without the free-stream turbulence in appearance. However the mass transport mechanism between them was basically different because of the formation of the ordinary separation line.

View full abstract

This study concerns experiments and analyses of the condensate flow and heat transfer in a dryer with a scraper in the case where the condensate film is thin. In the analysis using a model, the film thickness diverges at relatively slow rotations. A criterion for the critical condition on which the divergence commences is proposed. The analysis of the heat transfer is possible with a simpler model termed the solid film model, unless the divergence occurs. The experiments of the heat transmission are conducted with an acrylic resin cylinder and a stainless steel cylinder, and the flow patterns are observed. The experimental results on heat transmission agrees with the analytical one if about a 25 μm thick condensate film remains in spite of scraping.

View full abstract

The interruption of a fin surface causes renewed boundary layers starting from the leading edge of each interrupted fin. This results in boundary layer thinner in the average over entire length of fin surfaces compared to the uninterrupted case. Therefore, a higher degree of interruption is normally expected to lead to a higher degree of heat transfer. However, to realize this expectation actually, care must be taken of the geometrical allocation of each fin in an array so as to avoid other undesirable effects relating to the fin geometry. Numerical computation of flow and thermal fields have been made for two-dimensional models of two types of parallel louver fins and an offset fin. Heat transfer characteristics of parallel louver fins are discussed in comparison with those of an offset fin based upon the results obtained from the performed numerical computation. It is demonstrated that a cartain type of parallel louver fin has a higher heat transfer performance than that obtained with an offset fin.

View full abstract

Flow instabilities in a nitrogen evaporator are studied experimentally and analytically to investigate the characteristics of the cryogenic flow instabilities and to confirm that the previous stability analysis can be applied to them. The nitrogen evaporator consists of two parallel spiral tubes. Two types of periodical oscillations are observed, one is the oscillation in phase for two tubes and the other is out of phase. The period of he oscillation in phase is independent to the heat flux of the tube and is considered to be caused by boiling at the upstream of heating area. The oscillation out of phase is considered to be a density wave oscillation as its periocd is almost equal to the residence time and occurs in the region where the slope of pressre drop to mass flux is positive. Stability maps are obtained and the stability criteria show that the exit quality is smaller than 0.6. Linear stability analysis for he density wave oscillation using a drift flux model predicts the experimental results fairly well.

View full abstract

Flow instabilities in a liquid nitrogen evaporator consisting of parallel vertical channels are studied experimentally and analytically both with nucleate boiling in the inlet section, the flow pattern of which transfers from bubbly flow to annular flow and with film boiling, from inverted annular flow to dispersed flow. As the channels are heated by electrical current, i.e., the wall heat flux is constant, both flow pattern transitions are obtained depending on the initial wall temperature for the same experimental conditions of the heat flux, the mass flux and the exit pressure. Density wave oscillations out of phase in the parallel channels are observed in both flow patterns and compared. Experimental results indicate that the flow with film boiling at the inlet section is stable for higher exit quality and its periods of oscillation are shorter than those with nucleate boiling at the inlet section . Linear stability analysis for the density wave oscillation predicts well the experimental stability criteria with nucleate boiling at the inlet section by using the previous drift flux model and those with film boiling by using a drift flux model formulated by analytical results obtained with a turbulent boundary layer theory for the inverted annular flow.

View full abstract

The flow and heat transfer characteristics of regenerator materials were investigated in oscillating flow not only for a conventional wire net but also for new materials such as sponge metal which can be set up easily in the Stirling engine. The prediction of pressure drop loss was possible by use of the hydraulic diameter as the representative length defined by the friction factor and Reynolds number. The regenerator efficiency and heat transfer coefficient were obtained by measuring the periodic temperature variations at both ends of the regenerators. A new relationship between regenerator reheat loss and regenerator size was also found from the analysis of the experimental data. The effects of the wire diameter and the length of the regenerator on the indicated thermal efficiency was clarified.

View full abstract

Free convection in a stably stratified fluid between two vertical plates was investigated experimentally using a NaI solution. Initially, uniform free convection cells were generated after one side wall was heated and the other cooled. The uniform cells grew into two type of cells, uniform cells and nonuniform cells, depending on the heat flux at the wall and the vertical concentration gradient of NaI. The thickness of the nonuniform cells varied, and some of them were so thin that virtually no free convection occurred inside them. On the other hand, free convection occurred equally in each uniform cell. Therefore, the heat and mass transfer in the nonuniform cells was greatly reduced from that in the uniform cells. The mass transfer coefficient was correlated using the Rayleigh number based on the temperature difference of the walls and the initial vertical concentration gradient of NaI. The transfer characteristics in these layers wese also clarified.

View full abstract

Detailed temperature profiles in the twin valveless pulse combustor have been measured using a newly-designed coherent anti-Stokes Raman spectroscopy (CARS) to understand the fundamental pulse combustion processes for the engineering design. Time-averaged and ensemble-averaged temperatures in a combustion chamber were obtained from a constant 20Hz data sampling and a phase locked data sampling, respectively. The CARS successfully demonstrated near real-time, spatially and time-resolved, nonintrusive temperature measurement. Results show that very large temperature oscillations of up to 600 K exist near the wall upstream of the combustion chamber, while stable and homogeneous temperature regions were created in the rest of the combustion chamber. The maximum temperature was found to be 200 to 400 K lower than the adiabatic flame temperature. Also, comparison of one-cycle ensemble data between temperature and heat release rate gives a similar tendency, this indicating that the fuel-air mixing rate affects significantly the ignition delay time.

View full abstract

Numerical predictions are compared with the results of experimentation on turbulent mixing in swirling pipe flows with density nonhomogeneity induced by the mixing of He or CO₂ with air. Computations employ two kinds of stress/flux equation models : The density-weighted Favre-averaged form and the density-unweighted conventional time-averaged form. The calculation based on the former model can predict that the mixing of He with air will be more strongly suppressed than the mixing of Co₂, whereas the latter model fails to predict this. The possible interpretations of the interrelation between turbulent mixing, density nonhomogeneity and swirl-induced pressure gradient are presented.

View full abstract