The purpose of this study is to establish finite-difference solutions in use of high-order-accurate upwind scheme. In this paper, numerical analysis method based on a fractional step method was presented. This method used 3rd-order Adams-Bashforth method for convective term and Crank-Nicolson method for viscous term. For spatial derivatives in Navier-Stokes equations, 5th-order-accurate upwind scheme for convective term and 6th-order-accurate centered difference scheme for viscous term are applied. By this method, incompressible viscous flow in three-dimensional cubic cavity is computed at Reynolds numbers of 400, 1000 and 5000, with 41×41×41 grid points. In the analysis of parallel plate flow, the solutions of this investigation are in good agreement with the result of DNS data by Kim et al.. As a result of this analysis, it was made clear that high order numerical analysis for incompressible flows could be carried out by the method suggested in this study.
Three-dimensional unsteady blood flows through the artery with a small side branch were analyzed numerically using the finite element method. Our main attention was paid to the model that had a large branch angle (>90°) between the trunk and side branch. This arterial geometry is easily found in the actual arterial system. In this study, the arterial wall was assumed to be rigid and the non-Newtonian effect of blood was not taken into consideration. The calculated results are summarized as follows : 1) The branch angle does not have a strong effect on the flow rate through the side branch. 2) In the case of large branch angle, the still standing stagnation point appears near the outer wall of side branch. Moreover, this still standing stagnation point has an effect to increase the value of WSSG.
At 1st and 2nd reports, the Danamic Molecular Collision (DMC) model of diatomic molecules which is based on the molecular dynamics is constructed. The equilibrium state, transport coefficient at certain temperature, and the density profile of normal shock wave can be calculated well by this model. However, it becomes clear that this model can not calculate the property of the distribution of energy of the MD simulation, the temperature dependency of transport coefficient and the distance between density and rotational temperature profile of normal shock wave. For this reason, the DMC model is improved to calculate these values accurately in this report. First, we introduce the new potential parameter obtained from the potential which is based on the ab-initio calculation. Second, the collision cross section which is based on kinetic theory is introduced to calculate the temperature dependence of transport coefficient well.
In the 3rd report, we introduced an LJ potential parameter which is based on the potential obtained from ab-initio calculation and collision cross section from the Wang-Chang Uhlenbeck and Taxmas's theory and Molecular Dynamics (MD) calculation. In this 4th paper, we improved a Dynamic Molecular Collision (DMC) model to calculate the property of MD simulation better than the former one. To make sure of its validity we calculated equilibrium state, transport coefficient (viscosity coefficient, heat conductivity) at some temperatures and normal shock wave by the DSMC method using DMC model and compared these results with other theoretical and experimental results. Consequently, we found that the diatomic rarefied gas flows could be simulated very well using our model.
This study describes a flow structure in the hydraulic spool valve in low pressure. The spool valve is made of acrylic resin with the spool diameter of 24 mm. Three slit widths of 0.8 mm, 1.2 mm and 1.6 mm are employed in the experiment. Flow pattern was visualized by laser sheet. A flow entered into a chamber of the valve was initially disturbed, and gradually regulated as two dimensional flow in annulus between two concentric cylinders (d=13 mm and D=24 mm). In this region, streamlines initially placed at middle of the annulus were sucked toward the spool surface. Some streamlines located near the wall of housing were suddenly bent upstream of a slit because of decreasing pressure. Downstream of the slit a jet was mixed with rotating flow at upper side, and on the other hand a vortex was generated at bottom side by a jet which collided a bottom wall of housing. Three dimensional numerical analysis was performed by a finite-element method for a downstream model and a full model under conditions of steady and turbulence. Results of numerical analysis in a full model agree with experimental ones.
We describe the unsteady flow around two-dimensional circular cylinder bundles supported by the leaf spring in the fluid-elastic vibration. The experiment was carried out in a circuit-type wind tunnel having a 200 mm×200 mm working section of 2000 mm length at the Reynolds number Re of 6300. Two-dimensional circular cylinders with diameter D=20 mm were aligned at regular spaces S in the square array of three rows. The spacing ratio between two adjoining circular cylinders was constant as S/D=1.5. The time-mean velocity, turbulent intensities and Reynolds stress were measured by use of a laser Doppler velocimeter. The unsteady flow pattern around an oscillating cylinder was observed by the flow visualization in a water channel. Consequently, it was found that the occurrence of the fluid-elastic vibration is caused by the phenomenon of wake-switch for the circular cylinder bundles of three rows.
Many components of organized fluctuating pressures with different frequencies, which may be induced by the turbulent organized structures in the wake, have been extracted from the surface pressures on the finite circular cylinders with aspect rations from 1 to 12 at a Reynolds number of 3.7×104 using phase-locked averaging technique. The organized local and global fluctuating fluid forces acting on finite cylinders have been determined from these results. For cylinders with high aspect ratios, as large as more than 8, three different components of organized local fluctuating fluid forces act, i.e., tip component which appears near the tip exhibiting the largest amplitude, the midspan component as well as the root component which is similar to that for a two-dimensional cylinder. The lateral force and the rolling moment dominate in the organized global fluctuating fluid forces for many aspect ratios, but only a little organized fluctuating fluid force acts on the cylinders with aspect ratios from 5 to 6.
There are many works on the fluid flow around two circular cylinders arranged in tandem. An interesting phenomenon was found by one of the authors, Igarashi, in the range of L/d=1.2∼1.3, where L is the longitudinal distance between the axes of two cylinders of the same diameter. Namely, the Strouhal number decreases from 0.25 to 0.10 with an increase in the Reynolds number. In the present study, the behavior of the unique flow and the oscillating fluid forces acting on the two cylinders were investigated. The diameter of cylinder, d, was 80 and 100 mm, the distance L/d=1.2, the free stream velocity U ranged from 6 to 20 m/s. The Reynolds numbers were, therefore, 3.1×104∼1.3×105. The results obtained were as follows : The shear layers separated from the upstream cylinder reattach alternately on the downstream cylinder. A large lift force acts on the reattachment side of the downstream cylinder. There is a phase difference of 4/16 period of the vortex shedding in the oscillating lift of the two cylinders.
Prevention of vibration and acoustic resonance, originating in cross flow through the tube banks is important for a heat exchanger such as boiler. This paper is concerned with the excitation mechanism of velocity fluctuation in the tube banks which are arranged closely together. We measured the volocity fluctuation in a closely staggered arrangement. The ratio of Ion-gitudinal pitch to rod diameter of this arrangement was varied in a range of L/D=1.1 to 1.5, while the traverse spacing ratio is fixed at T/D=2.0. The range of Reynolds number is from 0.33×104 to 6.6×104. There are three dominant frequencies in the flow, caused by the different mechanisms : the fluctuation of flux which are flow rate fluctuations due to periodical change in the reattaching point, the change of the flow pattern in the separation region and the vortex shedding. The fluctuation of flux and the change of flow pattern have strong influence on the flow, and dominance in the two mechanisms changes with Reynolds number and the longitudinal spacing ratio.
The trajectories of a single bubble in tap water and turbine oil in a horizontal spiral tube, in which a swirling flow occurs, is investigated numerically using the Lagragian method. Bubbles in the range of 0.5 to 3.0 mm diameter were used in the simulation. The results show that the bubbles converged at a certain radius and the radius increases with an increase of the bubble diameter and viscosity. The formula for critical convergent radius is obtained using dimensional analysis.
An experiment was conducted in order to investigate the flow behavior for an air-water two-phase flow in a vertical pipe rotating with it's own axis. From flow visualizations, it is found that a unique flow pattern appears in the rotating two-phase flow as the bubbly slug flow which exists between the bubble flow and the slug flow modes, and it is revealed that these modes which are strongly influenced by the speed of pipe rotation have strong relation with the entering basic flow. Basic flow characteristics are much influenced by the rotation, followed by the reduction of the pressure drop and changes in the void fraction, depending upon the speed of pipe rotation and entering two-phase flow modes.
Annular jet has a larger contact area with the ambient than that of a round jet when the flow rate is same, and also the flow charcteristics can be controlled by coaxial round jet. The authors made clear already the mean and the fluctuating flow characteristics of annular air jet and their control by coaxial round air jet. In this paper the flow characteristics of gas-solid two-phase annular jet and their control using coaxial round air jet are discussed by the numerical and experimental analyses. The numerical analysis was carried out using k-ε turbulent model and two-way method in order to take accout the effects of dispersed phase (fine solid particles) on continuous phase (air). The turbulent charceteristics were evaluated by random simulation.
The opposed turbulent jets have wide applications and are seen in combustors and impact fluidic amplifiers etc., The flow fields employed by the two nozzles which one has very small diameter and the other has very large diameter are very different for the another cases. In this case, the small jet is perfectly covered by the other large one. As a result, the boundary conditions of the small jet are under the control of the large one. Momentum of the small jet is mainly diffused by turbulent diffusion on the mixing layer. Consequently, the diffusion mechanisms are quite different with the other cases. In this article, new understanding is presented for the flow fields of the opposed turbulent jets by measurements of the static pressure fields and turbulent flow fields.
The cold flow experiments are made in a two-dimensional dump diffuser. Flow behavior over a flame dome head and its reattachment to the casing wall are presented, when a symmetric or distorted flow is provided at the diffuser inlet. A Laser-Doppler Velocimeter was used for the velocity measurements. The effects of three types of flame tube depth and two types of vertical and inclined wall in a sudden expansion region on the diffuser performance are clarified. Both flame tube depth and inclination of the combustor wall are the dominant parameters affecting the diffuser performance of total pressure loss and flow rate distribution into the branched channels.
This paper describes the characteristics of pressure fluctuations in the shock wave/turbulent boundary layer interaction region induced by blunt fin placed normal to the wind tunnel wall. The effects of Mach numbers and fin configurations on the pressure fluctuations are also presented. Experiments were conducted at free-stream Mach numbers of 1.5 to 2.6 for blunt fins having sweepback angles of 0, 15, 30 and 45 degrees, with fin leading edge diameters of 4, 6, 8 and 10 mm, respectively. The results showed that the maximum pressure fluctuation levels near separation at given Mach numbers were a function of fin leading edge diameters and boundary layer thickness. It was also found that Strouhal numbers based on shock traveling frequency, shock traveling region length and free-stream velocity have a constant value of 0.020 for all blunt fin configurations.
To investigate the mechanism of drag reduction by dilute addition of polymer in turbulent flows, a direct numerical simulation (DNS) of turbulent channel flow including discrete-element models has been carried out. The influence of polymers is approximated by beads-spring-dashpot model. The number of grid point for fluid turbulence is 64×64×64 and the number of discrete-elements is 250000. The DNS result achieved 15% drag reduction. The viscosity represented by a dashpot, rather than the elongated elasticity by a spring, contributed to the drag reduction. Experimental evidence such as enhancement of streamwise fluctuation, reduction of Reynolds shear stress, increase of streak spacing in the near-wall turbulence was consistently reproduced in our simulation.
The interaction between the grid-generated turbulent flow and a 2-dimensional circular cylinder is investigated experimentally. Particular attention has been devoted to the effect of the velocity fluctuations induced by the cylinder wake vortices on the upstream turbulent field. The simultaneous velocity measurements at three different points around the cylinder (one in the upstream region and two in the near wake) have been performed by the I-type hot-wire probes. The velocity signals are analyzed by the phase average technique on the basis of the reference velocity signals in the near wake. It is found that the phase averaged velocity in the upstream region of the cylinder shows the large value when the vortices are growing up behind the cylinder. The fluctuation of the phase averaged velocity increases as the cylinder surface is approached and the largest value appears along the line of θ=90° near the cylinder surface. Some kinematic vortex-shedding model is suggested to estimate the intensity of the wake-induced velocity fluctuation. The spatial distribution of the calculated velocity r.m.s. fluctuation shows qualitatively good agreements with the experimental results.
The experiment was carried out under conditions that the temperature of the lower wall was kept constant and the range of Reynolds number was lower than 30000 (Rer=121∼507, Tr<2.0°C). About the fully developed channel flow of velocity and temperature, the authors investigated the effects of Reynolds number and heat flux. As a result, distributions of time mean velocity hold a universal law and distributions of time mean temperature also show universality, that is, they are independent of Reynolds number and heat flux. Particularly, the distribution of universal time mean temperature coincides with the case of pipe flow. And the distributions of velocity turbulence intensity are independent of Reynolds number, but are strongly affected by friction temperature. Therefore we have introduced the factor KY in putting results in order, which means the ratio of the velocity turbulence intensity for the heated steady flow to that for the non-heated in the viscous layer, and investigated relation of friction temperature with this factor. This factor relates only friction temperature and the empirical function is shown by the equation of KY=1+0.0854×Tr+0.1027×T2r. In the viscous layer, the skewness and flatness factors of velocity fluctuation are agree with nonheated steady flow and not affected by friction temperature. Those factors of temperature fluctuation have the same feature as the velocity fluctuation.
Measurements of time-mean velocity, turbulence intensities (u', v') and Reynolds shear stress in plane turbulent Couette flow have been made in the range of 3000≤2hUb/ν≤20000 (Ub, moving belt speed ; 2h, channel height). The main objective of this work is to study the effects of low-Reynolds number on the turbulent Couette flow. Particular importance is that its effect on the law of the wall can be estimated only from the Couette flow experiment. Low-Reynolds number effects are shown as follows ; additive constant of the law of the wall B, a coefficient of defect law and the maximum turbulence intensity u'/u* (u*, friction velocity) decrease as the Reynolds number decreases. From the skewness-and flatness-factors and four quadrant analysis of the fluctuating velocity, the turbulence structure of the Couette flow can be deduced. Except for close to the wall (y+≤12), where the structure is the same as that of the Poiseuille flow, the sweep event becomes more important than in the Poiseuille flow. The central region has a similar structure as a uniformly sheared flow
A hybrid grid method to solve the Navier-Stokes equations is applied to three-dimensional flowfields of crossing shock wave/boundary layer interaction caused by asymmetric double fin. The numerical method on hybrid unstructured grid is developed using a finite volume cell vertex scheme and the LU-SGS implicit time integration algorithm. Two kinds of one-equation turbulence models are implemented and evaluated for their accuracy by comparing with experiments and structured grid computations. The present numerical method is also applied to an inside flowfield of a real scramjet engine inlet configuration to demonstrate the capability of the method.
Effects of screens and two-dimensional contractions on a small open wind tunnel have been studied experimentally. Different mesh size of screens, number of screens, contraction ratios, nozzle length and nozzle shapes are investigated. Hot-wire anemometer is used for measuring the velocity profile and the turbulence intensity at the exit of the wind tunnel. It is found that the turbulence intensity decreases with the increase in the contraction ratio and number of screens. The turbulence intensity depends slightly on the mesh size of the screens. The short length nozzle makes the uniformity of the turbulence in the horizontal direction worse. The nozzle with the cosine curve can be regarded as better than the Batchelor-Shaw nozzle contour for the two-dimensional contraction.
Interaction phenomena between two turbulent spots were experimentally investigated in a zero pressure gradient laminar boundary layer. The turbulent spots were generated by issuing small air jets from two horizontally displaced holes on a flat plate. As the spots grew downstream, they merged into a larger single spot and the merged spot conformed almost the same outline to the superposition of the individual spots. The mutual spanwise intrusion into the other spot was restrained in a interaction region near the merged part, though the free wing tips on the other sides of each spot grew independently. Apparently, the interaction occurred only in the limited area narrower than the superposed region. There, both the positive and negative peaks of the ensemble-averaged velocity fluctuation of the merged spot increased than the single spot, and the calmed region was elongated while the normal growth was enhanced.
The vortex generator jet method as an active control technique of flow separation provides a time-varying control action to optimize performance under a wide range of flow conditions. The interaction between the jets and the freestream generates longitudinal vortices. The jets in this study were skewed at 90 deg to the freestream. The effect of jet pitch angle on suppressing flow separation was studied experimentally in three cases with the pitch angle set at 30, 45, and 60 deg. For the 60-deg case, longitudinal vortices move more rapidly apart from the lower wall than the other cases and therefore the 60-deg case is inferior to the others in the control of boundary layer separation. A pitch angle of 30 or 45 deg makes separation control more effective. In particular, for a pitch angle of 45 deg, the effect on suppressing flow separation persists further downstream.
It is very important in high enthalpy shock tunnels to predict heating conditions of the shock tube end where the tube wall is exposed to high temperature. The final goals of these studies are to establish methods to predict thermal conditions to which a tube end is subjected, and to develop thermal protection which will be suitable for high enthalpy shock tunnels. In the present report, a numerical flow analysis method combined with a heat conduction analysis of tube wall was developed to calculate the flow and wall terperature conditions in a shock tube. A numerical procedure was proposed which includes calculation of the mixing region of driver and test gases, and a friction factor model is presented which considers the change of visconsity under high terperatures. Experiments were carried out using the free-piston shock tunnel T5 of the California Institute of Technology. Comparison of calculated and experimental results of pressure histories in the shock tube showed that the present computer program has enough accuracy to calculate the flow field at the shock tube end.
In this paper, the generating process and the diminishing process of the nonuniformity regarding the strength and the shape of the shock wave front propagating through branched ducts were explored by numerical simulation. Computations were carried out by solving the two-dimensional compressible Navier-Stokes equations by using the total variation diminishing (TVD) scheme. Computations were performed for six types of branch (90° branch, 1/2 size 90° branch, T-shape (1/1, 1/2, 1/3 size) branch, 45° branch) and two incident shock Mach numbers (Ms=1.3, 2.0). The flow field were numerically visualized by the pressure and vorticity contours, velocity vectors, pressure distributions on the walls and time histories of the pressure contours on the walls. The mechanism of the generating process and the diminishing process of the nonuniformity of propagating shock front was clarified.
In this paper, the shock wave propagating through the plenum chamber with Borda's mouthpiece was studied by numerical simulation. Computations were carried out by solving the two-dimensional compressible Navier-Stokes equations by using the total variation diminishing (TVD) scheme. Computations were performed for five types of the plenum chamber with Borda's mouthpiece [coaxial mouthpiece, off axis mouthpiece, oblique edge mouthpiece, no mouthpiece (coaxis, off axis)] and two incident shock Mach numbers (Ms=1.5, 2.0). The flow fields were numerically visualized by the pressure and vorticity contours, velocity vectors and sonic lines. The pressure distributions and time histories of the pressure contours in the wall and the axis of the duct were also studied, and the attenuating process of the shock front and the diminishing process of the nonuniformity of the shock front were explored.
A flexible vibrating plate with distortion, fluid transport system of comparatively simple structure, accelerates surroundings fluid locally. However, it is not well known how the vibrating plate with deflection drives the fluid around it in the CFD field. Therefore, a practical finite-difference method that is able to handle such complicated phenomenon is desired strongly. In the present study we have not used the ALE method. which is often applied these problems. All terms of NS equation are assigned by the velocity which is observed from the grid on the moving mesh in this method. In addition, we used the new mesh generation technique because the solver needs quick response. As the mesh is moving with the elastic plate, we call this Elastic Mesh Generation (EMG) method. As the results, this solver which indicated good agreement between the experiment and the calculation can be used as a practical tool to predict the velocity field near the plate and the reacting force from the fluid to the plate.
Propagation of solitary pressure wave through a thin-walled collapsible tube filled with a viscous incompressible fluid is calculated numerically. One-dimensional continuity equation, equation of motion including a wall friction term, and the tube law obtained experimentally are used as the basic equations. The pressure wave-forms at three locations along the collapsible tube calculated by the present method are compared with those obtained by the experiment. It is shown that the tube law obtained experimentally needs to be modified to make the slope discontinuous at an onset point of the locally two-dimensional buckling of the collapsible tube. The viscous effect is evaluated using an equivalent Womersley number a for the solitary wave. Incorporation of the wall friction term with high accuracy is necessary in the case of small a in one-dimensional flow calculation. The propagation velocity can be approximately calculated using the Moens-Korteweg velocity corresponding to a tube having circular cross-section when the disturbance velocity is small.
The stability of alternate blade cavitation in flat plate cascades was examined by a linear analysis with a singularity method based on closed cavity model allowing the cavity length change. It was found that alternate blade cavitation was stable for the cascades with larger solidity and unstable for the cascades with smaller solidity. The equal length cavitation is stable only in the region of cavitation number larger than that where the alternate blade cavitation solution starts to separate from the equal length cavitation. This result suggests that the alternate blade cavitation could exist only for the cascades with higher solidity.
This study is intended to investigate the cavitation at the end surface of a cylinder when this is impulsively put into a longitudinal motion along its axis. The principle is that a longitudinal stress wave accompanying the longitudinal particle displacement is generated at one end of the cylindrical rod by an impact given at the other end ; the stress wave then propagates into the adjacent water as pressure fluctuations causing cavitation when the negative pressure reaches a certain value with a certain time duration. The results of the investigation can be summarized as follows (1) Cavitation is more susceptible with the air content in water. The pressures at the onset of cavitation increase with the air content. (2) Though the minimum pressures at the test surface decreases with increasing impact speed of the rod, the cavitation occurs at a pressure higher than the minimum and this incipient pressure is almost independent of the impact speed. (3) The incipient cavitation pressure can be negative when the air content in water is reduced from the saturated value.
When the valve in water supply pipes is closed suddenly, the flow stops and the momentum changes with the inertia of water flow. We call it as "Water hammering". A fundamental study of water hammer softening in household water supply pipes is carried out experimentally and theoretically. Three kinds of enlargement tubes of 80, 120, 160 mm in inner diameter with the same length of Φ200 mm and four kinds of the enlargement tubes of 120 mm inner diameter with five kinds of length of 3, 5, 10, 50, 200 mm were tested in order to reduce the pressure rise and the slope of the pressure rise. As a result, it was found that as the diameter of the enlarged part became greater, the reduction of the pressure rise became greater and the slope of the pressure rise became greater.
A numerical analysis is developed to the sound propagation of spinning acoustic mode in a non-uniform three-dimensional cylindrical ducts without flow. A linerized unsteady Euler equation for acoustic disturbance is solved by a fourth-order accurate finite difference scheme. The three-dimensional structures of the sound fields are obtained for several frequencies of sound source. The computed sound field is fully consistent with theoretical results with respect to cutoff frequencies. The present formulation for reflection and transmission coefficients shows effectiveness for the three-dimensional sound field.
The conformability of rated power output of the system to wind conditions is investigated theoretically with our simulation model on a wind turbine-generator system consisting of a Darrieus-Savonius hybrid wind turbine and a load alternator. As a computed result of the dynamic behavior of the systems which are operated at a constant tip speed ratio, it is clear that a large-scale system is effective with net extracting power in catching wind power at high velocities. However, an excessively large-scale systems do not always have an advantage over small-scale systems because the rotational speed allowed by the strength of the blade and excessively high velocities limits the power extraction. Regarding the effective power coefficient, it is clarified that a small-scale system is more effective than a large-scale one because of the dynamic characteristics of the system. Hence, it is shown that a conformable rated power output of the system exists for each wind condition.
The purpose of this report is to bring out the natural convection heat transfer characteristics of a finite vertical thick plate on the adiabatic floor, which is surrounded by the constant temperature wall. Numerical results were obtained for the Grashof number in a range of 0.1 to 105, the ratio of hot and cold wall temperature ranging from 0.05 to 1.0 and Plandtl number of 0.71. The effect of interaction on the natural convection fields were investigated. As a result, in case of small wall temperature ratio, when the Grashof number is large, convection fields of the hot wall side and the cold wall side can form each fields independently. On the other hand, as the Grashof number decreased, a fluid from hot wall side flows into convection filed on cold wall side. It was found that negative heat transfer coefficients at upper end of cold wall are calculated.
Heat transfer in a rotating rectangular duct was numerically simulated by using the fourth order finite difference method. In order to investigate the effects of the Coriolis force and the duct cross-sectional aspect ratio on the turbulence, the large eddy simulation was adopted with a dynamic subgrid scale model. In the computation, the aspect ratio and the rotation number were varied from 0.25 to 4.0 and from 0 to 5.0, respectively, while the turbulent Reynolds number was set to be a constant value, 350. In this aspect ratio range, both direct and indirect effects of the Coriolis force on the turbulence were seen, that is, the direct influence on the fluctuating velocity and the indirect influence through the modified streamwise velocity profile by the secondary flow. The relative degree of these influences depended on the aspect ratio. The Colburn's j factor showed the higher value for the larger aspect ratio cases because of the larger heat transfer enhancement on the pressure and side walls.
An experimental study was performed on the enhancement of steam absorption into a LiBr aqueous solution. The enhancement method proposed here is to set an absorption heat transfer plate facing downward to generate free convection within a liquid film falling over the plate. The experiment was conducted varying the tilt angle of the heat transfer plate in a range of 40 to 90 degrees, and the heat transfer coefficients were compared for both cases of the heat transfer plate facing upward and downward. It was found from the experimental results that the downward facing plate has higher heat transfer coefficient than the upward facing plate, which confirmed that free convection and mixing occur in the liquid film and enhance the absorption heat transfer in case of downward facing plate. The enhancement effect becomes remarkable as the tilt angle approaches horizontal, though the liquid film becomes thick and its thermal resistance increases.
Absorber of water/LiBr absorption chiller is made up of horizontal tube banks, and its improvement is needed to get higher COP. To clarify the absorption process in these tube banks, a single row of horizontal pipe column is studied experimentally and analytically. The analytical model proposed in previous paper is further changed to conservative form according to the method of Shyy. It is confirmed that the mass balance of water is kept exactly by the present procedure. As for total absorption rate for the pipe column, agreement between experiment and numerical calculations are fairly good except the cases of 8 to 10 pipes at lower flow rate. The roughed pipes with different knurls on their surface were also tested in the experiment and it was found that knurl height of 0.5 mm was the most effective in absorption enhancement.
This paper analytically treats of the heat-flow through the periodically varying thermal-resistance thin layer lain between conductors or the heat-flow between two periodically contacting conductors. During one cycle of the period, p, thermal-conductance of the thin layer temporally changes in a stepwise profile, and takes two different values, Rc and Rd over the time spans, φp and (1-φ) p, respectively. Analytical solution for the effective thermal-conductance, Re, of the thin layer was algebraically approximated using Rc, Rd, φ and the harmonic mean, Rab, of the characteristic thermal-conductances of the two conductors. Algebraical approximation was also made for the maximum temperature amplitude, Δθ*max, arising at the joint ends of conductors. This approximation suggests the following features of Δθ*max, Δθ*max takes the maximum value, (Δθ*max)max, at specified value of φ, φmax, φmax depends only single parameter Rc/Rd, and becomes large with increase of this parameter. On the other hand, (Δθ*max)max depends not only Rc/Rd but also Rd.
Heat removal of more than 10MW/m2 in heat flux has been required in high heat generation equipments in nuclear fusion reactor. In some condition of water subcooling and velocity, there appears an extraordinary high heat flux boiling in transition boiling region. This boiling regime is called micro-bubble emission boiling (MEB) because a lot of micro-bubbles are spouted from the heat transfer surface accompanying a huge sound. This study is aimed to obtain heat transfer performance of MEB in horizontal and vertical heated surfaces to parallel flow of subcooled water, comparing with CHF of this system. Three types of MEB with different heat transfer performance and bubble behavior are observed according to the flow velocity and liquid subcooling.
A mathematical model for heat production due to thermal excitation of conductive electrons and positive holes in semiconductor pn junction is derived and discussed. The model is applied to simulate the thermal runaway phenomena in power electronics semiconductor devices. Our discussion focuses especially on the modeling of unexpected huge current due to excessive temperature rise. Calculated dynamics of temperature distributions of silicon wafer while cooling performance decreases proved it possible that silicon wafer might be heated over its melting point in a few miliseconds. Our results indicate that if local hot spot arises in wafer, thermal excitation of intrinsic carries increases diffusion current of minor carriers and recombination current in depletion layer of pn junction. And it appears to be important that cooling performance should be uniform on the wafer to avoid the growth of hot spots and thermal runaway itself.
It is important that copper films fill the sub-micron trenches on silicon substrates in order to produce high-density electronic devices. In this paper, we numerically calculated changes of the copper film profiles in trenches on silicon substrates in the reflow process by 2 and 3-dimensional analyses. Calculation results from the model of surface diffusion in a thin layer on the film were similar to experimental results. The relationship between the change of the film profile and the geometric parameters were calculated. The results show that a uniform and thick initial film on the side walls of a trench is important to avoid film-bridge and void formation in the trench. In order to promote flow rate into trenches, a thick initial film on the bottom of a trench and narrow trenches are effective.
The compressed air energy storage gas turbine (CAES-G/T), an innovative method of meeting peak demand requirements of electric utilities is now on the installation in Japan, where surplus power is used to compress air and store it in a high-pressure underground circular horizontal cavern. Then it serves as a supply to help drive a gas turbine during periods when extra generating capacity is needed. The ultimate objective of the present study is to develop a methodology to analyze the unsteady state response of outlet air temperatures to a variation of the temperature inside and of the cavern. In the present work, an analytical model based on the two-dimensional laminar flow on the cross-section of the circular cylinder was developed to quantify the effect of the transient process occurring in the cavern and wall during injection, storage and release of the compressed air in the experimental circular cavern. It was found that the stratified temperature distribution was maintained in the cavern during compression and expansion periods. The wall temperature varied together with the variation of the air temperature with time, leading to the heat storage and/or release in the wall.
This paper deals with the heat transfer characteristics of a liquid-liquid direct contact operation in which a Perfluorocarbon (PFC) liquid is released in a hot water stream, a low-grade heat source such as urban sewage, for the purpose of heat recovery from it. The paper reports on a set of experiments in which a PFC liquid (1800 kg/m3 at 20°C) was continuously injected from a single, downward-facing nozzle into a slow, upward flow of hot water to be disintegrated into droplets descending in, and thereby heated from the water flow. The results of the experiments show how the size distribution and the translational motions of the droplets affect the overall coefficient for the water-flow-to-droplets heat transfer and also the temperature effectiveness for the droplets.
The Heat Recovery Steam Generator (HRSG) becomes more complex in accordance with the increase of gas turbine firing temperature. Recently, the triple stage steam pressure is one of the major system of HRSG. Such kind of HRSG requires a large amount of experimental parameter analysis to achieve the appropriate performance. This paper presents an deductive design method using pinch technology and linear programming, instead of the conventional parameter analysis. This method can be applied to the optimum heat recovery system design which minimize the total (energy and capital) cost. This method is also useful for investigation of the more complex heat recovery system, such as the steam cooled gas turbine system or the gas turbine fuel heating system.
This paper describes the numerical simulation of a GM refrigerator operating at about 4K. The simulation model is composed of a regenerator, an expansion space, and a cooling stage. The volumetric change of the expansion space and the actual thermophysical properties of the working fluid, helium are considered. The basic equations are made up of one-dimensional fluid equations and an energy equation of the regenerator material and the cooling stage. The fluid equations are expressed in the general coordinate system of which a coordinate axis moves with time to take account of the volumetric change of the expansion space. These basic equations are differentiated by using the TVD MacCormack method. In order to verify the simulation model, the theoretical validity of the calculation results was checked and calculation results were compared with results of experiments. As a result it was confirmed that the simulation model is appropriate.
The procedure to determine controlling parameters of pre-freezing and freezing processes of biological tissues are discussed. In the pre-freezing process, the concentration of cryoprotectant is raised stepwise to avoid osmotic stress. The rate of increase of the concentration of cryoprotectant can be chosen by considering the relaxation time of membrane permeability of a biological cell. On the other hand, during freezing process, the initial concentration of the cryoprotectant and the cooling rate must be determined to avoid osmotic stress and freeze injury. Optimum parameters can be obtained by estimating the normalized volume of the cell. The maximum possible size of a biological tissue to be cryopreserved can be estimated from the view point of the viability of its cells.
Low-temperature flames established on a flat-flame burner are investigated using rich diethyl ether/air mixtures to acquire a new knowledge of self-ignition processes. Temperature, chemical species and emission spectra are measured along with the burner axis. A series of cool flame, blue flame and a bright yellow column appears in a case of equivalence ratio 2.0. A dark zone demarcates between the blue flame and yellow column. Fuel enrichment causes a yellow column disappearance. Onset temperatures of cool and blue flame are 400 and 850 K respectively, which are independent to the equivalence ratios. Blue flame emission spectra contain an emission ranging from 720 to 810 nm in wave length, which can be estimated to be the H2O spectrum. Yellow column spectra are similar to that of black bodies; the emission probably from solid carbon particles therewith. The dark zone has similar near-infrared emission spectra and intensity to the blue flame though the visible emission is quite weak. The appearance of blue flame is closely related to an intensive decomposition of fuel, i.e., only when the decomposition reactions complete in the blue flame, the low-temperature oxidation process is raised to the final hot-flame stage.
The quenching effect due to flame stretch and heat loss was evaluated based on a HTA (Hyperbolic Tangent Approximation) theory for a counter premixed flame. The main results are as follows : (1) Reaction progress variable c at the stagnation point of the counter flow decreases with flame stretch and heat loss. The flame quench occurs under the condition of c=0.7 at the stagnation point. (2) Blow-off of a Bunsen burner occurs under the condition of g>(1.62-Nu(δ/d)2) (Su/δ) (g : velocity gradient, δ : thickness of laminar flame, Su : laminar flame speed, Nu : Nusselt number, d : diameter of burner). This condition was in good agreement with experimental data. (3) Flash back occurs under the condition of g<0.697 (Su/δ).
This paper reports the suppression of soot emission from acetylene diffusion flame by applying electric field. The effects of applied voltage, polarity and spacing of electrodes on soot emission were investigated experimentally. The results showed that the shape of flame changed and the soot emission decreased with increasing the applied voltage. The polarity of applied voltage influenced to the shape of flame and the soot emission. When positive voltage was applied to the nozzle electrode, the flame length became short and the width at the flame tip was spread. And, more than 90% of soot emission was suppressed at over 200 kV/m of the intensity of electric field. The surface temperature of flame increased with increasing the applied voltage. Thus, it was considered that the increase of flame surface temperature promoted the soot oxidation. When negative voltage was applied to the nozzle electrode, on the contrary, only the 70% of soot emission was suppressed, because the flame temperature was comparable to that for non-applied voltage.
In the previous paper, the authors clarified that an impingement spray have the slip flow region near the impingement point on the wall. Then, behavior of a diesel spray impinging on an inclined wall was modeled to make an empirical formula of impingement spray tip penetration. To introduce a new model of spray growth behavior, uniform velocity penetration of the spray growth behavior in the slip region were framed and included into the traditional formula about the free spray. The momentum conservation theory was applied to the spray of after slipping. The growth behavior of an impingement spray was expressed as the function of wall distance and inclined wall angle. Spray path length calcurated from this formula showed qualititaively well agreement with an actual measurement.