In the complete lock-in phenomenon of vortex shedding with small oscillation of a body at average natural vortex shedding frequency, the flow behind the body is considered to exhibit no irregular characteristics. However, even under the complete lock-in condition, the flow behind the body, in particular the vortex formation length, was found to exhibit irregular characteristics. In this study, the time variation of the vortex formation length was investigated by image processing of flow visualization, Also the cross correlations of the vortex formation lengths between the upper and lower separated shear layers were investigated. From this study, it was clarified that the breakdown of vortex formation is associated with large vortex formation length.

抄録全体を表示

In order to find useful methods for estimating the rate of severe cavitation erosion arising in a typical centrifugal pump at high specific speeds such as N_s=300(m, m³/min, rpm), systematic erosion tests are carried out under various test conditions using three-dimensional impellers made of aluminum alloy. The following three effective and simple methods for estimating the erosion rate are presented. (i)Observation of the initial aspects of the eroded surface, (ii)Analysis of the intensity of the vibrating accelerations, and(iii)Examination of the relationship between the erosion rate and the position of maximum erosion rate.

抄録全体を表示

A Vortex Generator Jet (VGJ) in a turbulent boundary layer has been investigated, since it has a significant potential for the active control of boundary layer stall. This paper describes the pulsed VGJ through a single hole which generates periodic longitudinal vortices in the boundary layer. Phase-averaged streamwise velocity and wall shear stress were measured by a flow sensor with a high response. Under some pulsed injection conditions, higher velocity is observed in the region downstream of the injection hole and close to the wall at the end of injection. As a result, phase-averaged wall shear stress has a peak at the end of injection. The optimum injection condition is also obtained.

抄録全体を表示

A Vortex Generator Jet (VGJ) in a turbulent boundary layer has been investigated, since it has a significant potential for the active control of boundary layer stall. This paper describes the double VGJs under pulsed injection which generate periodic longitudinal vortices in the boundary layer. Phase-averaged wall shear stress was measured by a wall flow sensor with a high response. The important flow parameters of double jets are their arrangement and phase difference of injection. In common flow down arrangement with common phase, phase-averaged wall shear stress has a higher peak at the end of injection than that of a single jet.

抄録全体を表示

An experimental investigation is conducted on supersonic internal flows over a rectangular cavity at mach number M=1.83. In this paper, the basic characteristics of the cavity flow fields and the effect of cavity dimensions (the length to depth ratio) are discussed. Time-averaged and unsteady pressures are measured and the flows are visualized by the schlieren and the oil flow techniques. The basic characteristics of the supersonic cavity flow, namely three-dimension-ality and the shock/shear layer system, are recognized. A longer cavity has a large circulation flow and a pair of small three-dimensional vortices due to the separation on the side walls. A structure of the flow is highly three-dimensional. The shorter cavity has only one vortex and the flow is essentially two-dimensional. The unsteady pressure data reveals the dominant frequencies of the cavity flow phenomena. In this investigation, the features of the basic cavity flow fields are clarified.

抄録全体を表示

At present, the DSMC method is the best scheme to analyze rarefied gas flows. However, polyatomic inelastic collisions which have complicated energy transfer between translational and internal degree of freedom cannot be analyzed accurately. In this paper, a model for energy exchange between translational and rotational degrees of freedom is constructed using the results of MD simulation. In this 1st report, we describe how to simulate the collision of diatomic molecules by the MD method and the results of the simulation, We find that the energy transfer is distributed without a characteristic relation to the direction and orientation of the rotational vector, and that both the energy transfer and collisional cross section strongly depend on the initial translational and rotational energy.

抄録全体を表示

In the 1st report, we made some calculations to obtain data in order to construct a model for collision of diatomic molecules. Consequently, we found that the energy transfer is distributed without a characteristic relation to the direction and orientation of the rotational vector, and that both the energy transfer and collisional cross section strongly depend on the initial translational and rotational energy In this 2nd paper, we constructed a collision model for diatomic molecules using these data. To make sure of its validity we calculated equilibrium state, viscosity coefficient, heat conductivity diffusion coefficient and normal shock wave by the DSMC method using our model and compared these results with other theoretical and experimental results. Consequently, we found that the diatomic rarefied gas flows can be simulated very well using our model.

抄録全体を表示

The scattering of an oxygen gas molecule from a graphite surface is simulated using the molecular dynamics method. The numerical simulation is compared with experimental results obtained by scattering a supersonic molecular beam of O₂ from a clean graphite surface in a UHV chamber. The angular distributions of scattered O₂ of both results show good agreement. The translational and rotational energy loss of the gas molecule at the first collision is also analyzed. The rotational energy loss is most affected by the initial rotational energy, and the translational energy loss is most affected by the initial normal translational energy. Both energy losses are also affected by energy transfer. Comparison with a monatomic gas molecule shows that the core size and the potential well are important factors for the scattering distribution and the energy loss of the gas molecule.

抄録全体を表示

The rotational spectra observed in O₂-LIPF(laser-induced predissociative fluorescence) experiments using an ArF^* excimer laser are theoretically investigated. First, the simulation code of the Schumann-Runge bands is developed and then the general characteristics of the bands are discussed from the viewpoint of the suitable wavelength for O₂-LIPF. Although the tuning range of the ArF^* excimer laser is not the best wavelength regime for the O₂-LIPF, it is shown that the range of the excimer laser is adequate to determine rotational temperatures and that the bands (4, 0), (7, 1), (10, 2), (11, 2)and(14, 3) are observed to be intense there. Finally, the actual spectrum, considering the effects of the laser linewidth and the predissociation linewidth, is discussed and it is predicated that the (14, 3) band is most affected by those linewidth effects among the dominant bands above.

抄録全体を表示

Prediction of fuel rod vibration in BWRs is important for designing a new fuel assembly An experimental study was performed to obtain the rod vibration characteristics for a 2×2 rod bundle in parallel boiling two-phase flow Vibration characteristics of fuel rod assembly were obtained under various conditions, e.g., system pressures of 0.98∼7.16 MPa, flow rates of 430∼1990 kg/m²s, inlet subcools of 33∼125 kJ/kg and outlet qualities of 0∼25%. The main results obtained from the high-temperature and high-pressure tests are as follows. The characteristics of vibration direction and response magnitude of a fuel rod were found to be random, and the amplitude was reduced to 1/5∼1/10 when the system pressure was increased from 1 MPa to 7 MPa. The effect of boiling on the rod vibration obtained to reduce the rod vibration response because of an increase of damping effect due to boiling.

抄録全体を表示

A precise investigation of the stability of the boiling two-phase flow of magnetic fluid under a nonuniform magnetic field is conducted in relation to the development of a new energy conversion system. First, governing equations of boiling two-phase flow based on the unsteady thermal nonequilibrium two-fluid model are presented and analytically solved using a linearization method. Then the analytical results of magnetic stabilization are inspected experimentally by flow visualization and image processing techniques using an experimental apparatus composed of a small test loop. From the theoretical and experimental studies on the stability of boiling two-phase flow of magnetic fluid, the stabilization of two-phase flow is achieved due to the magnetic force of the fluid and appropriate superficial gas-phase velocity. Also, it is clarified both theoretically and experimentally that the axial magnetic field more effectively stabilizes the two-phase magnetic fluid flow than the transverse magnetic field.

抄録全体を表示

The purpose of the present paper is to understand how boiling two-phase flow behaves in a vertical circular tube for the case of small mass velocity and small heat flux. Flow visualization study was made to investigate the effects of the mass velocity and heat flux on boiling two-phase flow in the tube. It has been found that the flow pattern and/or flow pattern transition is influenced by the inner diameter of the tube, and that the height of the dry-out point depends on the mass velocity, heat flux and inner diameter of the tube.

抄録全体を表示

The quantitative measurement of an image obtained by neutron radiography technique was proposed in order to accurately measure the void fraction of a two-phase flow in a metallic duct. It was shown experimentally that the spatial distribution of the dark-current component was homogeneous and the temporal variation could not be ignored. Since the scattered neutrons falling on the image converter could be homogenized by setting the test section at a distance from the converter, it was clarified that the corrections for dark-current and scattered neutrons could be represented by an offset value. It was proposed that the offset value could be determined by using the total macroscopic cross section of the material (Σ-scaling method). By comparing the calculated void fractions with the measured ones obtained by simulating the known void profile using a standard test section, the void fraction could be measured by this method within 2% error. The measurement error was estimated to be up to 10% when no corrections for scattered neutrons were made or no arbitrary offset value was used.

抄録全体を表示

Cerebral artery aneurysms often form in the circle of Willis, most frequently in the asterior communication artery. This artery forms a junction between two confluent arteries and two bifurcating arteries. In the present study, the flow field around the anterior communicating artery is simulated by two confluent tubes, two bifurcating tubes, and junctional region connecting to four tabes. The wall shear stress and the velocity profile around the anterior communicating artery in laminar steady flow were experimentally studied using an electrochemical method and a laser Doppler velocimeter, respectively. The flow around the junction is found to be very unstable and the wall shear stress to vary greatly around the apex.

抄録全体を表示

The purpose of this study was to clarify the flow pattern inside a drum that was suddenly stopped from a constant speed. Ten cases were observed by flow visualization using light sheet methods and by measurement of the time history of the velocity using LDV. The experimental results are expressed in terms of two nondimensional parameters: π₁, which is determined by the viscosity of the liquid, the speed of rotation, and the size of the drum ; andπ₂, which indicates the time required to attain e^-1 of a constant speed The relationship between these two parameters is given by a single formula. Prior to each experiment, π₁ was set and the time required for the drum to come to rest was estimated from the time required to attain e^-1 of the constant speed using this relationship andπ₂.

抄録全体を表示

The pressure distributions of pulsatile flow through a two-dimensional channel bounded by a fixed wall and an oscillating wall are analyzed numerically. In this calculation, a coordinate transformation method is used to analyze the moving boundary problem. The analyzed results are summarized as follows: (1) An increase in the amplitude of the wall oscillation leads to an increase in additional pressure loss due to the oscillation of wall. (2) The wall oscillation does not lead to a decrease in the pressure loss under the present calculation conditions. (3) The phase lag between the flow pulsation and wall oscillation has a strong effect on the additional pressure loss. However, this effect becomes weak with a decrease in the oscillation period of the wall.

抄録全体を表示

Experimental investigation is carried out in order to determine the effect of the cutting parameters such as abrasive feed rate, nozzle discharge pressure, nozzle traversing speed, nozzle size and material of the specimen on the striated roughness of the surface in abrasive waterjet cutting. It is shown that when the combination of nozzle traversing speed and nozzle discharge pressure is chosen so that the maximum kerf depth is kept constant, a smooth surface can be obtained by the combination of slow nozzle moving speed and low discharge pressure. It is also shown that the wavelength of striation ranges mainly from l to 1.5 mm, regardless of the cutting parameters.

抄録全体を表示

Experimental studies are carried out to clarify the effects of a polymer additive on the structure of a submerged water jet and the damage process of a metallic material. Tap water and an aqueous solution of polymer additive are pressurized up to 147 MPa and injected into still water at atmospheric pressure through a nozzle of 0.50 mm in diameter. Specimens of aluminum alloy are used for erosion tests. The structure of the submerged water jet is analyzed from instantaneous photographs. It is found that the structure of the submerged water jet is affected greatly by the addition of a polymer additive. Furthermore, the addition of a polymer additive improves the cutting and drilling capacity of the submerged water jet.

抄録全体を表示

Conditional sampling measurements were made to reveal the turbulence structures of a steady round jet and a pulsating round jet with low pulsation frequency and relatively large velocity amplitude using a hot-wire anemometer and a laser Doppler velocimeter. Turbulent motions were classified into four distinct categories : ejection, sweep, outward interaction and inward interaction. The fraction of time the velocity signals were in the respective categories and the contributions of each motion to the Reynolds shear stress and turbulence kinetic energy were determined. Ejection and sweep were found to be responsible for turbulence production. The correlation coefficient between the axial and the radial velocity fluctuations showed that large-scale ordered motions existed in the jet. There was little difference between the turbulent structures of the steady and pulsating jets.

抄録全体を表示

The effects of a passive control boundary layer applied around the throat on a transonic diffuser were investigated experimentally by wall static pressure measurements and by schlieren optical observations. The experiments were conducted using three kinds of the diffuser walls, one was a solid wall and the other two were porous with a 30 mm or 50 mm-long cavity underneath enabling the flow around the shock wave to circulate through the porous wall. The results show that the Mach waves normal to the flow were observed when diffusers were almost choked and that the pressure fluctuations in a transonic diffuser were greatly reduced by passive control. According to the frequency analysis, the frequency range attenuated by passive control is between about 700 Hz and l kHz regardless of the length of the cavities.

抄録全体を表示

Self-excited oscillation of an underexpanded jet issuing from a circular nozzle and impinging perpendicularly on a flat plate is analyzed numerically using a TVD scheme. With some nozzle-plate spacings, simple patterns of the oscillation of the flow fields are obtained numerically, and their pressure histories at the center of the plate also show simple wave patterns with a certain amplitude of the pressure. Their amplitudes increase with increase in the nozzle-plate spacing, and a dominant frequency of about 20-30 kHz is obtained by FFT analysis of those pressure histories. The numerical results agree with the experimental data. Such self-excited oscillation seem to be related to the noise of the jet. The amplitude of the pressure history is considered to depend on the area of the compressive region upstream of the plate shock. It is also found that the pressure wave which is generated near the impinging region travels toward the nozzle wall in the air surrounding the jet, which causes the whole jet to oscillate.

抄録全体を表示

A two-dimensional finite element method for an incompressible two-fluid model is developed in order to enhance the applicability of the two-fluid model to the analyses in arbitrarily shaped channels. The solution algorithm is based on the SMAC method for incompressible single-phase flows. Quadrilateral element with four nodes is employed. The velocities and volumetric fractions of both phases are defined on the nodes. The pressure is defined at a centroid of each element and assumed to be constant within the element. The Galerkin method is applied to the finite element formulations. Air-water two-phase flow around a circular cylinder is analyzed by the finite element method. The calculated distributions of the volumetric fraction of the gas-phase show good agreement with measurements.

抄録全体を表示

Three typical mechanical heart valves (Starr-Edwards, Bjork-Shiley c-c and St. Jude Medical) have been tested in the mitral position under pulsatile flow conditions, The test program included measurements of velocity and turbulent stresses at six downstream locations. The study was carried out using a sophisticated cardiac simulator in conjunction with a highly sensitive two-component laser Doppler anemometer system. The continuous monitoring of parametric time histories revealed useful details about the complex flow and also helped established location and time of the peak parameter values. Based on the experimental data, the following general conclusions can be drawn: (i) all 3 valve designs generated elevated turbulent stresses during acceleration and peak flow phases, presenting a possibility of thromboembolism, and perhaps, hemolysis; (ii) valve configuration and valve orientation significantly affect the flow characteristics; (iii) the Starr-Edwards valve with relatively a small opening area shows lower turbulent stresses; and (iv) the bileaflet design of the St. Jude Medical valve is also associated with lower turbulent stresses, as compared to the Bjork-Shiley c-c valve.

抄録全体を表示

This study deals numerically with the aerodynamic sound radiated from the flow over a circular cylinder at a Reynolds number of l0³. The standard acoustic analogy method is applied. The incompressible time-dependent two-dimensional flow around the cylinder is calculated by means of the finite difference method, then the variables are substituted into the approximate analytic solution of a nonhomogeneous wave equation. The result of far-field sound calculation using a solution which Howe expresses as a volume integral form is compared with that using another one which Curle expresses as a surface integral form. Good agreement is obtained between them. Howe's equation is used to investigate the sound source distribution in the flow field. The sign of the pressure fluctuation radiated from a small sound source region is opposite to those radiated from the neighboring ones. Most of the pressure fluctuations interfere and vanish as a consequence of volume integration. The remaider is observed as far-field sound. It is shown that there are some regions which contribute more significantly to the sound generation than the other regions on average in time.

抄録全体を表示

The purpose of this paper is to estimate the renormalization group theory for turbulence developed by Yakhot and Orszag (1986) and reformulated by Yakhot and Smith (1992). We examine their basic theory for the Navier-Stokes equations, and the transport equations for the kinetic energy K and its dissipation rate ε. It becomes evident that their method is not related to Wilson's renormalization group theory. Their K-ε model is not directly obtained by the renormalization group method. They evaluated the Kolmogorov constant by setting ε=0 and ε=4 in the same equations. Furthermore, the constants in the K-ε model are invalid because of the same problem.

抄録全体を表示

An experimental study of the developing turbulent flow in an elliptical duct with an aspect ratio of 2 : 1 was carried out using a laser-Doppler anemometer. It is pointed out from numerical analysis that the intensity of the secondary flow of the second kind shows a maximum value before the flow becomes a fully developed turbulent flow. Therefore, special attention was paid to the measurement of Reynolds stress and the secondary flow of the second kind in developing turbulent flow in order to confirm the numerical prediction. Since it is very difficult to measure the secondary flow of the second kind because of its low intensity, we utilized a laser-Doppler anemometer which can measure flow without disturbing the flow field As a result of this study on developing turbulent flow, the experimental distribution of the secondary flow is found to coincide with the numerical prediction ; therefore, the features of developing turbulent flow include the highest intensity of the secondary flow. At the same time, a clear distinction of the distribution of five of the six Reynolds stresses was made, showing their characteristic aspects. The experimental results, except for the normal stress in the vertical direction, agree with numerical predictions without much discrepancy.

抄録全体を表示

A numerical calculation is performed on the turbulent channel flows with uniform injection using five different low-Reynolds-number κ-εmodels. Published experimental data on the mean velocity profiles, the Reynolds shear stress, and the friction velocity are used to assess the performance of the model calculations. Applicabilities of these models to turbulent channel flows with injection are investigated. Fof flow with injection, the κ-εmodel proposed by Launder and Sharma gives good agreement with the experimental data. Also, the effects of the inlet velocity and the channel width on those flows are examined using the model which provides most accurate prediction.

抄録全体を表示

The finite-element method is universally applicable to fluid flow analysis in complex geometries. In finite element analysis, ordinary iterative methods of solving the Poisson equation repuire large amounts of computation time to satisfy the equation of continuity. In order to reduce the computation time, reduction of iteration in the Poisson equation is required. Increase of iteration occurs particularly when using an irregular mesh and in the analysis of three-dimensional problem. In this paper, we apply a modified weighting area method for GSMAC iteration and examine a new method for high speed calculation of the Poisson equation. The Navier-Stokes equation is written in rotational form using only a mass matrix and discrete del operator as the element coefficient matrices in order to reduce memory storage in the computer. The lid-driven cavity flow problem is chosen to verify the validity of the present numerical method.

抄録全体を表示

Generally speaking, the finite-element method in computational fluid dynamics is a universal numerical method, but in the computation of a flow phenomenon with unstructured grids, it takes much CPU time to construct computational grids. Therefore it is important to develop an analytical formulation for the purpose of reducing the time required to construct computational grids. In the present paper, in order to overcome the defect of the finite-element method, we propose a new mixed- element method using the discrete del operator defined as an element average of the gradient of the shape function in discrete space. The analytical expression of the discrete del operator for the mixed-element method is a vector in two or three dimensions, and becomes a nonmemorizing method. Furthermore when we use the discrete del operator, the natural description of programing becomes objective and compact.

抄録全体を表示

Generally speaking, finite-element methods in computational fluid dynamics are universal, but uneconomical. In the present paper, in order to overcome this defect in FEM, we propose a new method using the discrete del operator which is a coordinate-free differential operator in discrete space. This operator in discrete space is defined as an element average of the gradient of the shape function, and it has three characteristics: orthogonality, identity and symmetry. Furthermore, the discrete del operator is useful in nonmemorizing and in easy coding. Since the analytical expression of the discrete del operator is a vector in two or three dimensions, the natural description of programing becomes objective and compact, which is more understandable for nonspecialists of CFD.

抄録全体を表示

The implicit time-marching finite difference formulations of the Euler and Navier-Stokes equations using the upwind TVD scheme for two-dimensional compressible flows are applied to the flow field in supersonic through-flow fans. The results of the present Euler and Navier-Stokes equations are used to investigate the flow field for various geometrical and flow parameters. It is shown that the difference between the Euler calculation and the Navier-Stokes calculation is conspicuous in the region near shock waves or near the trailing-edge of the blades. Prediction of overall performance, however, is not susceptible to the two calculation models since the difference is limited to narrow region and is of small magnitude, unless the separation and reverse flow regions are generated.

抄録全体を表示

A dynamic load balancing technique is applied for solving the Euler equations on networked engineering workstations Message passing is handled by PVM(Parallel Virtual Machine) software. A domain decomposition method is employed to parallelize the algorithm. In this load balancing technique, the convergence criterion is evaluated on each subdomain. When the criterion is satisfied on a subdomain, the computational domain other than that subdomain is redecomposed on each processor. This procedure is performed once for supersonic flow, and several times for transonic flow. In the present approach, the total load of the calculation of two-dimensional Euler equations solvers is reduced by 48% for the test case of supersonic flow in a duct and by 29% for transonic flow.

抄録全体を表示

It was recognized 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. Though it has been noted that the aerodynamic force has some effect on the vibration of the train, few studies have been conducted and little knowledge has been gained. In this study, to clarify the aerodynamic effect on the train, a three-dimensional unsteady incompressible Navier-Stokes simulation was executed. Flows around the train both in an open section and in a tunnel section were simulated. Unsteady flow separation on the rear nose, which results in fluctuation of the yawing moment of the tail car, was obtained. In the tunnel section it was proven that the tunnel wall causes flow separation to be asymmetric and the expansion of the effective cross-sectional area of the flow between the tunnel wall and the train along the rear nose increases the pressure fluctuation. These factors result in the greater vibration in the tunnel section. We also presented a modified shape of the rear nose, which restrains flow separation and reduces the yawing moment fluctuation.

抄録全体を表示

There are two mechanisms which give rise to turbulent noise radiated from a counter- rotating axial flow fan. 0ne is the strong turbulence induced by the front rotor during the incoming main flow to the rear rotor, which, in turn, induces random force fluctuation on the rear rotor blade. The other is the shedding of vortices from the trailing edge of the both front and rear rotors. In this paper, we take these two mechanisms into consideration to theoretically estimate the turbulent noise level. The estimated values agree well with the measured levels of turbulent noise generated from the counter-rotating fans. It is theoretically clarified that the noise generated from the rear rotor is higher than that from the front rotor. The noise due to the turbulence during the incoming main flow is nearly equal to that due to the vortex shedding from the trailing edge of the rotor.

抄録全体を表示

From the standpoint of power conservation, bondgraphs can represent various systems by a unified approach., even which are composed of mechanical, hydraulic and electric componenst. In this paper, a torque converter is analized by bondgraphs following a set of four first-order differential equations describing torque converter charavteristics. A new bondgraph model of torque converter is presented for establishing a bondgraph of a stator. In this new bondgraph, the converter range and the coupling range are represented clearly. In the converter range, stator is not rotating, but tranmits the torque to the tranmission case, so the bondgraph must have a gyrator element for the stator. On the other hand, in the coupling range, the stator are free to rotate, so the bond of stator is omitted. By use of this new bondgraph, the transient condition from converter range to coupling range becomes possible to estimate. Dynamic characteristics of this torque converter can obtained easily by a bondgraph simulation in comparison with the usual simulation method.

抄録全体を表示

The effects of wall material and surface condition on the behavior of an initial boiling bubble of R113 subjected to transient heating is investigated using a heater with a large heat capacity. The behavior of the initial bubble is closely related to the premature transition to film boiling of liquids with high wettability. An initial bubble, which is peculiarly shaped like a 'straw hat' and leads to premature transition in saturated liquid nitrogen as reported in a previous report, appears also on the heated wall with large heat capacity and grows rapidly to cover the entire wall surface. The initial bubble is found to be a coalesced bubble into which small bubbles activated in succession along the heated surface are rolled, from the observation using a high-speed video camera. The growth rate of the initial bubble along the heated surface is not greatly affected by the thermal conductivity of wall material but is affected markedly by the surface roughness.

抄録全体を表示

A generalized numerical method for radiation heat transfer of absorbing, emitting and scattering media and specular and/or diffuse surfaces with three-dimensional arbitrary configurations, REM², is proposed. The ray emission model for various radiation elements is expressed by polyhedrons and polygons. Arbitrary thermal conditions can be specified for each radiation element, and generalized radiation transfer can be achieved regardless of surface and volume elements by introducing a new definition of view factors. The accuracy of the present method is verified using simple configurations. A cubic participating medium with a spherical cavity covered with specular and/or diffuse surfaces is analyzed as an example of an arbitrary configuration. The temperature distribution shows good accuracy in the case of a small number, such as 45, of rays emitted from each element compared with the Monte Carlo method.

抄録全体を表示

An experiment of heat transfer on a wall in a laminar flow was performed by using a wall jet. The wall jet was generated by a flow control plate which was inserted into the wall. The heat transfer coefficients were measured by Mach-Zehnder interferometer. The flow pattern and velocity were measured by a smoke-wire method and laser doppler velocimeter, respectively. The height of plates (h) was varied from 2 to 8 mm. The spaces between wall and plate (c) were varied from 0 to 7.6mm. The inclined angles of plate (θ)were varied from 60 deg to -60 deg at the constant front projected height. The following results were obtained. The large plate height gave a large local heat transfer coefficient. The local heat transfer coefficients were enhanced about 7-fold of whithout plate at h=8 mm, θ=30 deg, and c/(c+h)=0.15. The optimum inclined angle of control plate for maximum local heat transfer was 30 deg, and the optimum wall jet generator angle for heat quality was -30 or -45 deg.

抄録全体を表示

The heat transfer coefficients and the pressure distributions were measured downstream of the backward facing step in which the heat transfer was actively or passively enhanced by following procedure. (1)The jet was ejected perpendicularly to the main flow from the portion on the opposite wall of the step. (2) The air flow was sucked from the recirculating zone. (3) A slat was inserted with an inclination to the main flow. These cause the main flow to curve sharply, and then the recirculating zone to shorten, so it is easy to control the lengths of the reattachment point, and consequently, the heat transfer. The thermal performance, η defined as the ratio of the heat transfer coefficient at the recirculating region with these modifications to that of a normal region, was calculated for equivalent pumping power, and it was found that η was larger than unity for the enhanced systems, except in the case of the discharge jet. It is effective to apply the enhancement of the heat transfer by active or passive methods to the suddenly expanding tunnel.

抄録全体を表示

This paper presents three-dimensional simulations of an unsteady separated and reattached flow and the heat transfer over a blunt flat plate. A numerical analysis of the Navier-Stokes equations and energy equation is carried out using the finite difference method. The present results of three-dimensional calculations are compared with the previous results of two-dimensional ones and show an improvement of accuracy of the flow characteristics, demonstrating the importance of three- dimensional simulations for nominally two-dimensional flow. It is clarified from the numerical results that the separated shear layer becomes unstable and forms spanwise vortices. These vortices become three-dimensional and are shed from the reattachment flow region. The shed vortices possess hairpin-like structure. These large-scale vortex structures have great effects upon the heat transfer in the separated, reattached and redeveloping flow regions.

抄録全体を表示

This paper concerns the condensation of carbon dioxide and its heat transfer on a vertical surface in the subcritical region. In this region, physical properties of carbon dioxide show marked temperature and pressure dependence. In particular, as the pressure increases and approaches the critical pressure, the difference in the properties of liquid and vapor, and surface tension approach zero. Therefore, condensate flow from natural convection decreases and configuration of condensation may differ from that at lower pressure. In this study, the configuration of condensation is observed and at the same time, condensation heat transfer is obtained by measuring the volume of condensate. Heat transfer is discussed in connection with the configuration of condensation.

抄録全体を表示

Numerical calculations have been carried out to simulate comlex non-Newtonian fluid motion within a porous medium. A collection of square rods placed regularly in a region of infinite extent has been proposed as a two-dimensional numerical model of porous structure. The numerical simulation results obtained at a microscopic level with the power-law non-Newtonian model are processed to extract the macroscopic hydrodynamic characteristics in terms of the Darcian velocity and intrinsic average pressure. The macroscopic results in terms of the apparent permeability agree well with the results of the semi-empirical formula proposed by Christopher and Middleman.

抄録全体を表示

A measurement method of an unsteady temperature field by using a Mach-Zehnder interferometer and an image-processing technique was presented. In this study, the results measured by this method were compared with the calculated results under Reynolds numbers(Re= UL/v) ranging to less than 350 in the steady state. The unsteady characteristics of the local Nusselt numbers (Nu_χ=αχL/λ) around an elastically vibrating plate were investigated under the conditions that the reduced frequency (N=1/Ro) ranged from 4.8 to 8.6 and the amplitude of vibration was kept small relative to the plate-length. In this study it was clarified that the distributions of the local Nusselt numbers were observed flutteringly, and that the phase of the time averaged Nusselt number is delayed compared to the phase of the velocity at the trailing edge. These results indicate that the enhancement of the time-averaged heat transfer coefficient around the vibrating plate depended on the Rossby number.

抄録全体を表示

In this work, using Green's function method, the temperature wave solution predicted by the hyperbolic heat equation is developed for a finite medium exposed to a laser heat flux with an actual temporal profile, whose power has a finite growing and decreasing time. Using the analytical solution, the temperature response and the propagation of the temperature wave due to such heat pulses are investigated for different pulse durations, thicknesses of the medium, and energy absorption depthes. The possibility of exhibiting the temperature wave is discussed for a thin film under a short heat pulse, and the method for measuring the relaxation time is predicted.

抄録全体を表示

Numerical analysis has been performed on bubble growth and deformation under an electric field in order to elucidate the mechanisms of boiling heat transfer enhancement by EHD (Electro- Hydrodynamic) effects. Transient Navier-Stokes and Maxwell's equations were solved simultaneously for liquid and vapor phases in a two-dimensional cylindrical co-ordinate system making use of the VOF (Volume of Fluid) method. Bubble growth in liquid R113 under atmospheric pressure has been simulated. First, elongation of a single bubble under uniform electric field is simulated and the final shapes of the bubble are found to be in good agreement with Garton's analytical and experimental results. Second, the bubble deformation process under non-uniform electric field was simulated. A bubble initially attached to the lower electrode starts to deform and finally detaches from the lower electrode. The shape of bubble depends on the intensity of the electric field. The behavior of bubbles, the velocity vectors, and the contours of electric field are also shown and compared with experiment.

抄録全体を表示

As a result of recent progress in electronic equipment and devices, the power dissipation of LSI chips has tended to increase Because of this trend, it has become more important to improve the heat transfer characteristic of LSI packages. Methods already used to enhance heat transfer in manufactured packages involve adding a heat spreader or using heat radiation leadframes. In this paper, we propose two kinds of thermally improved packaging designs. One has molded plastic fins on top of the surface. The other has leadframe fins that extend from the heat radiation leadframes. We designed these thermal structures using a previously reported thermal resistance analysis of LSI packages, and then built trial models. We measured the thermal resistance of these models suspended in a wind tunnel. According to the measurements, thermal resistance of the package with plastic fins is about 34% lower than that of a finless package, and resistance of the package with leadframe fins is about 20% lower than that of a package with only radiation leads.

抄録全体を表示

This study analytically investigated the subcooling effect of injected water on falling water limitation in countercurrent two-phase flow (CCFL) in vertical channels, by applying a new model of momentum balances for both liquid and gas phases to the entire length of the channel. The subcooling effect of injected water on CCFL, which is one of the dominant parameters, had been clarified neither analytically nor experimentally because the CCFL phenomena is so complicated thermodynamically. As a result of the present study, it was clarified that the analytical model proposed here could give good predictions of the existing data on the subcooling effect of experiments simulating the performance of emergency core cooling water injection during a loss-of-coolant accident in pressurized light-water reactors.

抄録全体を表示

This paper describes a semi-empirical study on the development of a prediction algorithm of thermal resistance in the case of pin-fin heat sinks for LSI packages using impingement cooling. The inlet orifice was relatively large and was placed above the center of the heat sink. A physical model suitable for heat sinks with pin-fin arrays was considered based on flow visualization results. According to this model, the flow field was divided into five sections : (i) the top of the fin beneath the inlet orifice, (ii) the region under the inlet orifice, where the flow was parallel to the pin-fins, (iii) the fin base area, (iv) the region where the flow was perpendicular to the pin-fins excluding the region covered the bottom of the inlet orifice, (v) the top of the fin, except the region under the inlet orifice. Among them, heat transfer in the parts (i), (iii) and (v) was found to be neglective small, and it was not considered here. Thermal resistance values of the heat sink with pin-fin arrays were predicted for a variety of orifice diameters, gaps, pin-fin diameters, pin-fin heights and number of pin-fins. These values agree with the experimental data to within ±30% error.

抄録全体を表示

This study describes the transient temperature distribution in a cooling apparatus for high-power semiconductor devices used to drive motors of electric rolling stocks. The cooling apparatus is composed of heat pipes. In the model for simulation, we substituted solid elements for heat pipes, and determined their thermal properties by experiment. Consequently, the heat transfer rate of heat pipes can be obtained by a heat conduction analysis. Calculations show that when heat generation in the device changes, the temperature of cooling apparatus changes more slowly than that of the devices. A comparison between calculations and experiments confirms the accuracy of the modeling and prediction method.

抄録全体を表示

Experimental and analytical studies of the ice formation process with longitudinal fins in an isothermally cooled parallel-plate channel with convective water flow have been performed. The fins made of copper were soldered onto the cold wall parallel to the water flow direction. The resulting thickness of ice formed over the fins was measured under a steady-state condition in the ranges of Re_H = 710∼3060 and cooling temperature ratio of θ= 0.67∼5.13. An analytical method for ice formation associated with large fin spacing with convective water flow is proposed. In the present analysis, ice formation with "heat conductor plates", which are not physically soldered onto the cold wall, is also examined. The effects of spacing, thickness, height, and thermal conductivity of heat conductor plates on the ice volume fraction are investigated.

抄録全体を表示