The aim of the present work was to analyze the generation of vorticity on the surface of bodies. The fluid flow examined is viscous incompressible fluid and the local solution near the surface of bodies is considered. The solutions of the two-dimensional Navier-Stokes equation are constructed by means of a boundary layer equation for three initial vorticity fields, ω₀ (x, y)=-ω₀ (x, y), ω₀ (x, y)=ω₀ (x, y) and ω (x, y)=0 for y<0, where ω₀ is the initial vorticity and (x, y) is in the local coordinate system (x is tangential and y is normal). The results of the present analysis show that the vorticity generated on the boundary is determined by the slip velocity in these three initial cases for very small time lapse. This implies that the vorticity generated at the boundary is not dependent on the flow field models in numerical calculations. The present local solutions are applicable to the three-dimensional local flow near the boundary and the same conclusion is easily derived except flows near singular points such as separation and stagnation points.

View full abstract

We investigated the laminar-turbulent transition of a spherical Couette flow for clearance ratio β=0.14 at which velocity fluctuation disappears. We found that the root mean square value for V_φ shows complicated evolution for distribution in the radial direction, while the root mean square value for V_∂ shows a relatively simple one, where V_∂ and V_φ are the fluctuating velocity components of the azimuthal and meridian directions, respectively. Calculating the correlation dimension and drawing the Poincare section, it is revealed that the flow field traces a scenario as follows : steady state-> periodic state-> quasi-periodic state-> chaos-> periodic state-> steady state (the disappearance of velocity fluctuation) -> periodic state-> chaos. It is also shown that in the quasi-periodic state, the first return map becomes irreversible as the Reynolds number increases.

View full abstract

A numerical analysis of developing turbulent flow in a diffuser with S-shaped centerline at a Reynolds number of 4×10⁴ has been carried out. The diffuser exit-to-inlet area ratio was 1.5, the inlet cross section was square, 40mm×40mm, and was expanded on both curved surfaces to an exitplane section of 40mm×60mm. The S-shaped curvature was formed from two 22.5 degree bends of 280mm mean radius of curvature. Straight ducts 0.3m and 2m long were attached to the inlet and outlet planes of the S-shaped diffuser, respectively. In the calculation, an algebraic Reynolds stress model was adopted for predicting anisotropic turbulence precisely, and a boundary-fitted coordinate system was introduced for coordinate transformation. Calculation results for mean velocity were compared with the experimental data. In addition, the difference in flow between the S-shaped diffuser and the S-shaped duct is presented with calculation results. Using by the present method, we found that the flow at the inlet shows a "core" flow located near the inner wall of the first bend, and a region of low velocity fluid which has accumulated at the exit of the diffuser was observed. It was also found that the secondary flow patter of the S-shaped diffuser was similar to that of the S-shaped duct in the first bend but differed from it in the second bend.

View full abstract

A numerical analysis has been carried out for developing turbulent flow in a diffuser with S-shaped centerline at a Reynolds number of 4×10⁴. The diffuser exit-to-inlet area ratio was 1.5 and the inler cross section was square, 40mm×40mm, and expanded on both curved surfaces to an enitplane section of 40mm×60mm. The S-shaped curvature was formed from two 22.5 degree bends of 280mm mean radius of curvature. Straight ducts 0.3m and 2m long are attached to the inlet and outlet planes of the S-shaped diffuser, respectively. In the calculation, an algebraic Reynolds stress model was adopted in order to predict anisotropic turbulence precisely, and a boundary-fitted coordinate system was introduced as the method of coordinate transformation. Special attention is paid to the comparison of Reynolds stresses between calculation results and experimental data. In addition, the difference in turbulent structure between S-shaped diffuser and S-shaped duct is presented by invariant analysis using the calculation results. The present method can be used accurate prediction of the characteristic features of normal and shear stresses. Invariant maps of the S-shaped diffuser show that the turbulent structure transforms anisotropic turbulent flow into a two-dimensional turbulent flow. Moreover, the calculation results suggent that the turbulence level of the S-shaped diffuser increases more actively than does that of the S-shaped duct.

View full abstract

One of the authors has previously reported the behavior of the fluid dynamic drag of a twodimensional bluff body cut from a circular cylinder. The volume cut from the cylinder is equal to d/2 (1-cosθ_s), where d and θ_s are the cylinder diameter and the cutting angle (in the case of a circular cylinder, θ_s=0 deg), respectively. The drag coefficient takes a small value at around θ_s=53 deg when Re>Re_c (=critical Reynolds number) compared with that of θ_s=0 deg. In this paper, the flow characteristics around the cylinder, the dependency of Reynolds number and the effects of the cylinder's length on the flow direction are discussed, using four different sections of the cylinder in the range of Re=2.0×10⁴∼5.0×10⁴.

View full abstract

A new hyperbolic grid generation method is presented and developed for two-dimensional space in this paper. The method uses the backward-Euler implicit scheme in the marching direction, while a MUSCL-like upwind finite difference scheme with a minmod limiter is introduced in the other direction. Thus, the present hyperbolic grid generation method requires no further artificial damping terms, although a conventional hyperbolic grid generation method of a central difference scheme requires some artificial terms to dampen the oscillations. Body-fitted grids around a NACA0012 airfoil and a complicated U-shape body are generated. The hyperbolic grid generation method presented can be expected to produce excellent orthogonal property of grids. To investigate this, a criterion to estimate the orthogonal property of generated grids is introduced. The orthogonal property of the grids around the airfoil is demonstrated. The results show that the present method generates grids of good orthogonal property.

View full abstract

A globally fourth order non-oscillatory centred scheme is proposed for solving the compressible Euler equations. The present scheme can bring out stable solutions without multi-stage time integration, and calculation time is considerably reduced in comparison with MUSCL type schemes. As a numerical experiment for the present scheme, a computational result for a shock dynamics problem is obtained in comparison with some conventional schemes. The present scheme is useful for effcient computation for problems required higher time-space accuracy such as complex nonstationary flow field including weak discontinuities.

View full abstract

Active control of jets exhausting into a two dimensional channel with insulated suction near the wall during monitoring with a neural network is described. The aim of this work is to simulate the possibility of controlling the flow pattern with a flip-flop phenomenon by suitable suction. In this case, output of a neural network is estimated satisfactorily based on 3 flow patterns from the teaching data. The periodic flow pattern at Reynolds number Re=1.0×10³ is reduced by a less than 10% suction rate for primary jets. Then stable flow emerges as a result of feedback gain of the neural network. Since the control with suction does not have to be continuous, the power cost is reduced. The nonperiodic case at Re=1.0×10⁴ is hard to control, but we achieved good control by use of a suitable suction rate with the neural network.

View full abstract

The forces acting on and the flow patterns around a rigid bluff body submerged in oscillatory flow were studied. The force coefficients of a circular cylinder were measured and compared with those of other studies in a U-tube water tank to verify the experimental set-up. Then, measurements of both in-line and transverse forces of a square cylinder at an angle of attack of 0° were made in a relatively wide range of Keulegan-Carpenter numbers from 1 to 90. For the circular and square cylinders, the force signals were spectrally analyzed to get predominant frequencies. Flow visualization was also employed to confirm the correspondence between flow patterns and force coefficients. It is found that, for the present models of bluff bodies, several classes of flow patterns appear within the measured KC number range and the vortex motions closely corresponds to variations of the inline and transverse forces.

View full abstract

The motivation for this research was to obtain information on the separation characteristics associated with the phenomenon of stall. This separation data will ultimately be utilized as part of an optimum control system to reduce the negative effects associated with stall. Surface shear stress measurements were carried out using MEMS-based sensors on a two-dimensional wing with a Wortmann FX 63-137 airfoil section (chord Reynolds numbers ranged from 1×10⁵ to 6×10⁵). Computational airfoil performance from an airfoil analysis code was compared with the results of measurements. The wall shear stress distribution was measured by five micro hot-film sensors, which were flush mounted on the 2-D wing over the frontal 50% of the chord at various spanwise locations. Hysteresis characteristics were determined from time averaged and rms shear stress values as well as from airfoil performance curves.

View full abstract

The 3-D flow field relative to a rotating paddle blade is acquired by synchronizing LDV measurements with the output of a shaft-mounted encoder. This makes it possible to capture the impeller stream velocity field, including the details of the discharge flow region behind the blades. Mean velocities ensemble-averaged between a specific impeller blade pair (over 90°) show the extent of periodicity and the passage and structure of the discharge flow region. Pumping capacities are determined at both sides of impeller and tip of impeller, and it is shown that the radial jet is a consequence of fluid entrainment into the discharge flow region. Mean velocities taken in a rotating frame of reference reveal a significant periodic variation near the impeller than those taken in a fixed frame.

View full abstract

Buoyancy-driven helium-air exchange flow through double small openings was investigated. The double openings with rectangular cross section were settled side by side, in order to evaluate the interaction between the flow through the double opening. In the exchange flow experiment, the exchange flow rate and the flow pattern around the openings were simultaneously measured. Two flow patterns were formed even in the same boundary condition, i.e., (a) uncaptured flow pattern with larger exchange flow rate, and (b) captured flow pattern with smaller exchange flow rate. The entrainment effect caused the two flow patterns. In order to clarify the mechanism, characteristics of the entrainment was evaluated in the forced flow experiment. Only (b) flow pattern was observed in forced flow experiment. The relationship between the entrainment ratio a and flow rate q were measured for the flow patters (b). A new correlation for the exchange flow rate was proposed with considering the entrainment. With applying the a-q relationship for flow pattern (b) to the correlation, the exchange flow rate was predicted. It agreed with the measured value better than the conventional prediction without consideration of the entrainment.

View full abstract

The breakup conditions of a liquid droplet due to airstreams are investigated in detail. Experiments are conducted using a horizontal air suction wind tunnel. Uniformly sized liquid droplets produced by longitudinal vibration of a nozzle and uniform drops from a nozzle are used. The shuttering processes of liquid droplets are observed using a stroboscope and photographed by a still camera. The typical critical breakup pattern of a droplet in this experiment is of a bag type. The critical relative velocity increases as the diameter of a droplet decreases. The critical Weber number is not constant as stated in Hinze's theory, but increases as the droplet diameter decreases. The critical Weber number in the horizontal air suction wind tunnel is larger than that in the vertical air suction wind tunnel, and in the airstreams induced by a shock wave. The transition breakup zone which expands with decresing droplet size is also examined precisely. The relationships between droplet diameter and, for example, the relative velocity difference the rate of relative velocity difference, are shown quantitatively.

View full abstract

We present the characteristics of the critical pressure and discharge coefficient on the twod-imensional sonic nozzles at a low Reynolds number. The geometric shape of the nozzle is based on the ISO 9300 sonic-Venturi nozzle. The two-dimensional sonic nozzles of five different width were used in order to get low Reynolds numbers. The static pressure distribution along the wall of the nozzle and the behavior of the shock wave in the nozzle are shown. The discharge coefficients were obtained by the two critical sonic nozzles method to prove the mechanism of the reduction of the critical pressure ratio at a low Reynolds number. When the throat width was 2.5mm, the boundary layer near the throat was separated and, consequently, the recovery of pressure by the diffuser was not sufficient. The recovery of pressure was not improved with increasing the upstream pressure. The discharge coefficient of the 2.5-mm-wide nozzle was substantially decreased by the separation of the boundary layer, and the 20-mm-wide nozzle also decreased due to the formation of the corner vortex at the entrance of the two-dimensional nozzle. The discharge coefficient on the sonic nozzle is governed by two parameter ; α and β. At the three-dimensional sonic nozzle, a depends on the velocity profile at the throat and β is affected by the thickness of the boundary layer. On the contrary, at the two-dimensional sonic nozzle, the value of a changes markedly with changing throat width, but the value of β is constant since β is not affected by the throat width.

View full abstract

Steady flowfields resulting from slot injection at the surface of a flat plate in a freestream with a Mach number of 3.75 and unit Reynolds number of 1.65×10⁷m^-1 are simulated, using the compressible mass-averaged Navier Stokes equations and five typical turbulence models. The flowfields with various total pressure ratios of injection to freestream are computed, and comparisons are made with experimental data in terms of the surface total pressure distribution, the length of the upstream separation region, and the height of the Mach surface (i.e. penetration height). Two characteristics of model performances are clarified. Though the upstream separation length and the penetration height are differently predicted by each turbulence model, the correlations between them are nearly same. A Reynolds stress model can reproduce the upstream separation region more reasonably than a k-ε model does, especially for high injection pressure cases.

View full abstract

To improve the performance of a disk MHD generator, it is suggested to install inlet swirl vanes to introduce azimuthal velocity component in the supersonic inert working gas flow at the inlet of the generator channel. But the present proposed swirl vanes cause relatively large total pressure loss and non uniform flow. This paper first analyses the flow field around the swirl vanes and the mechanism of the fluid dynamical loss, then proposes a new asymmetric supersonic nozzle design for swirl vanes aiming at achieving the azimuthally uniform swirling velocity field with low pressure loss. The performance of these newly designed swirl vanes using asymmetric nozzles is confirmed by wind-tunnel experiments and two-dimensional numerical simulation.

View full abstract

This paper presents the experimental results of rotordynamic forces on an open-type centrifugal impeller in whirling motion. For the whirling open-type impeller, the variation of the tip clearance produces the rotordynamic force. It was found that, [1] destabilizing force occures at small whirling/rotation ratio throughout all of the flow range, and [2] At smaller flow rate the magnitude of the force changes dramatically at a whirling/rotation ratio. The forces estimated from unsteady pressure distribution on the casing and that estimated from the pressure difference across the blades are compared with the measured rotordynamic forces.

View full abstract

Since a rotating bucket of Pelton turbines intermittently penetrates into a free water jet, the confined sheet flow with free surface on the bucket is unsteady in space-time. In order to analyze the unsteady flow quasi-steadily, an animated cartoon method is proposed. Assuming the relative inflow to the bucket to be steady within the respective frame of the cartoon, the streak patches of the unsteady flow are computed by integrating the relative flow accelerations due to the constraint, lateral pressure difference and gravity, as well as the apparent centrifugal and Coriolis accelerations, considering the thickness of the water sheet under the atmospheric environment.

View full abstract

A new type of hydrocyclone with an internal perforated cylindrical tube is developed and its performance is experimentally examined for solid-liquid separation. Two types of the inner cylinder, whose void fractions of the perforated surface are 31% and 67%, respectively, are tested. It is found that the inner perforated cylinder removes the air core which is generally formed in the conventional hydrocyclone. It is also shown that the hydrocyclone with the inner perforated cylinder achieves better separation efficiency and lower pressure loss than the conventional hydrocyclone. The larger void fraction of the perforated surace elicits on higher efficiency.

View full abstract

Transient natural convection in water around a cooled horizontal cylinder placed in a rectangular enclosure has been studied. The governing equations were solved by a finite difference method, and the flow structure and temperature distributions have been predicted. The purpose of the present study is to examine the effects of the position of the cylinder and the initial temperature of water on the cooling process of water. The initial water temperature was varied at 4, 6, 8 and 12°C, and timewise variations of the temperature field and the velocity field as well as the mean Nusselt number on the cylinder surface were compared. The changes in the position of the cylinder and the initial water temperature largely affect fluid flows due to the density inversion of water. Complicated flow patterns are observed for initial water temperatures larger than 4°C. Therefore, the cooling rate of water is largely changed with the change in the position of the cylinder for initial water temperatures larger than 4°C.

View full abstract

Fluctuating temperature has been measured using a laser Rayleigh scattering technique along a centerline of turbulent stagnating opposed jets (cold jet velocity : 2m/s) with different temperatures up to about 400°C. The temperature signals show that the temperature fluctuation is characterized by spike type fluctuations with a sharp increase or decrease in temperature between hot and cold jets. This indicates that the temperature fluctuation is mainly attributed to the large-scale fluctuations of the boundary between the hot and cold jets. Profiles of mean temperature and rms value of fluctuating temperature can be normalized using a similarity parameter based on the bulk strain rate, characteristic intensity and scale of the turbulence when the temperature difference is less than 200°C.

View full abstract

In a compact closed two-phase loop thermosyphon which consists of a multi-radiator and a refrigerant bath with multiple power modules mounted vertically the heat radiation performance deteriorates as the refrigerant bath becomes thinner. We have found that this is because the return condensate is blocked by the backflow of the bubble in the return path when the heat load decreases and increases sharply. We have also improved the heat radiation performance by providing an adiabatic path between the boiling area and the return path.

View full abstract

For clarification of the cause and effect of fluid type on the generation of the Karman vortex street due to cooling of a cylinder at a low Reynolds number where the Karman vortex street does not occur in an isothermal wake, the two-dimensional, laminar, time-dependent continuity equation, Navier-Stokes equations with the buoyant term of Boussinesq's approximation, and energy equation are solved numerically for a circular cylinder wake with an upward freestream of mercury, air, and water. The cause is the generation of the wake vorticity which does not occur in an isothermal wake, and a stable arrangement of vorticity. With a decrease in the Prandtl number of the fluid, the Karman vortex street is generated easily by cooling of a cylinder, and has lower frequency, lower velocity of the vortex movement, and larger scale of vortex spiral than the isothermal Karman vortex.

View full abstract

We propose a numerical solution for the adjoint operator of a forced convection heat transfer problem to evaluate mean heat transfer characteristics under arbitrary thermal conditions. Using the numerical solutions of the adjoint problems under Dirichlet and Neumann conditions, both of which can be computed using a conventional CFD code, the influence function of the local surface temperature on the total heat transfer and that of the local surface heat flux on the mean surface temperature are obtained. As a result, the total heat fluxes for arbitrary surface temperature distributions and the mean surface temperatures for arbitrary surface heat flux distributions can be calculations using these influence functions. The influence functions for a circular cylinder and for an in-line square rod array are presented.

View full abstract

An experimental study is carried out on the characteristics and critical heat flux of ice accretion along a horizontal fine wire immersed in a cold air stream with water spray. The air stream velocity and temperature range from 3 to 10m/s and from -5 to -15°C, respectively and the average drop diameter of the water spray is about 0.14mm. To determine the effect of wire diameter on ice accretion, three diameters, namely, 0.5, 0.8 and 1.0mm are selected. The growth process of ice accretion along the wire over time is photographed by a video camera. The critical heat flux is determined from the profile of the variation of wire temperature with loading electric power on the wire. It is found that the effect of air srream velocity on the ice accretion along the wire becomes significant as the air stream temperature decreases and the critical heat flux shows a linear increase with an increase in the air stream temperature and velocity within the conditions of the present experiments. The effect of wire diameter on the critical heat flux, however, is not evident in this study.

View full abstract

Ice formation phenomena at the interface between cooled liquid and ethylene glycol solution are examined experimentally. It is found that the crystal ice formed at the interface removes from the interface by itself due to buoyancy and drifts away upward the solution, which suggests the possibility of "liquid-like ice" formation by this method. It is shown that the ice growth rate is suppressed with increasing the velocity of solution flow due to the increase of heat transfer from the solution to the interface. It is also shown that the analytical method for ice formation rate proposed in the present study is good to predict the experimental results.

View full abstract

A dynamic equation of state at the solid-liquid interface for a liquid-saturated "solid-liquid porous media" system has been derived from a thermodynamic study modeling simultaneous heat and mass transfer through porous media from the liquid supply front to the solid-liquid interface, as the very first step in building a mathematical model which may macroscopically describe a freezing expansion phenomenon with mass transfer in saturated porous media. It has been revealed that the derived dynamic equation of state generalizes the so-called generalized Clausius-Clapeyron equation, i.e. the static equation of state for porous media without mass transfer, and that mechanical energy produced at the solid-liquid interface is equal to mechanical energy dissipation due to the mass transfer.

View full abstract

The mixing effect deteriorates performance in a stratified heat storage tank. In the present study, the effect of characteristic factors during the charging process was analyzed, based on the analytical solution in the previous study in which a strict boundary condition was applied and a mixing effect was included. Also, the usefulness of the solution was verified and the method of application to a real system was examined by comparing it to other experimental results. It was concluded that the depth of the perfectly mixed region is dominant unless the Peclet number is very small.

View full abstract

Absorption refrigerator can be driven by waste heat. In particular, the double-effect type absorption refrigerator, which is highly efficient, is drawing a great deal of attention as a waste heat recovery system. This research refers to the double-effect absorption refrigerator driven by steam, assuming waste heat from the fuel cells is applied. In the start up of the double-effect absorption refrigerator, the solution temperature is almost equal to the temperature of the air. Due to the large temperature difference between the solution and the heat sources, the solution is over heated. In the worst case, it will be crystallized. Additionally, some solution is circulated due to the existence of a pressure difference between the heat exchangers. If the solution pump is started before and adequate pressure difference is obtained, the absorption refrigerator cannot be started normally. To investigate these problems, the simulation model is constructed. An experiment is conducted to investigate the performance of this model. As a result, the validity of this model is confirmed and the detailed start-up characteristics are clarified.

View full abstract

Studies were carried out to develop a high-performance ceramic heat exchanger for ultra high temperatures using a fluidized bed. Experiments were performed using a heat exchanger with a ceramic tube bundle immersed in the fluidized bed (maximum bed temperature was 1100°C). Although the cooling air temperature increased from 50°C to more than 1000°C, fluidization remained stable and the bed temperature uniform. A heat transfer coefficient of outer and inner side of ceramic tubes was obtained, and the effects of the bed temperature, fluidization velocity, cooling air temperature and cooling air velocity on the heat transfer coefficient were evaluated. The results show that the heat transfer coefficient for a fluidized bed was larger than that for single phase flow. This was due not only to the renewal effect by particles but also to radiation from the particles. The performance of the heat exchanger was also evaluated using the coefficient of overall heat transmission and temperature efficiency.

View full abstract

Pressure drop and heat transfer for flow-boiling of water in small-diameter tubes under atmospheric pressure were discussed. The results of experiments were compared with those of several existing correlations and models. The experiments were carried out using atmospheric-pressure water in tubes with inner diameter D ranging from 2.0 to 6.0mm, heated length L from 4.0 to 680.0mm, inlet water subcooling ΔT_in from 70 to 90K, and mass velocity G from 100 to 10170kg/ (m²·s). The highest heat flux q attained was 33MW/m². Interesting phenomena were observed in small-diameter tubes with a very long or very short heated length. The most suitable correlations for predicting the pressure drop and heat transfer in small-diameter tubes were discussed.

View full abstract

In the present study, the solidification and melting of water were investigated experimentally for the case of a vertical heat transfer plate with pin fins. In the experiment, temperature distributions, ice and water volume fractions, and heat flux changes were measured and the flow patterns in the water were observed for examination of the phase change process. In the solidification, the phase change rate increased monotonously with increasing number of fins. In the melting, the temperature distribution in water showed a uniformity caused by natural convection based on the density change of water. The contribution of the natural convection to the melting was examined based on the relationship between the modified Nusselt number and the Rayleigh number.

View full abstract

The objectives of the present study were to perform a numerical analysis of the solidification and melting of water around a vertical heat transfer plate with pin fins by means of a finite-difference method under a quasisteady-state assumption. The characteristics of phase change, such as the shape of the solid-liquid interface, the temperature distribution and the liquid velocity, were considered for different numbers of fins, and these analytical results were compared with the experimental results. We found that the heat conduction underwent a considerable increase when the number of fins was large during the solidification, but in the range near the heat transfer plate natural convection between the fins was confirmed for every number during the melting and small-scale natural convection around the pin fins was defected.

View full abstract

A method to solve inverse radiative property value problems is developed to obtain two-dimensional distribution of absorption coefficient in combustion chambers or furnaces from the profiles of the temperature and the radiative flux of the surrounding walls and of the gas temperature. In the analysis, initial guess of the absorption coefficients of each gas element are given at first, and by solving the forward problem, the wall radiative flux distribution corresponding to the given wall and gas temperature distributions is obtained. The initial guess of the absorption coefficients is corrected so as that the difference between the calculated and the given values of the wall radiative flux becomes small by using the successive linearlization method. The exchange factors required to solve the forward problem are obtained by the Monte Carlo method. In the present study, an algorithm is proposed to push the time-consuming Monte Calro process out of the convergence loop in which the values of the absorption coefficients are estimated and the Monte Carlo procedure is carried out only one time before the convergence process begins.

View full abstract

Structural investigation of absorbents at the molecular level is of very importance for the knowledge of their solubility. The EXAFS measurements of LiBr aqueous solutions with different concentration and with another cation, such as potassium ion, K⁺ or calcium ion, Ca²⁺, were carried out, X-ray absorption spectra at the bromine K edge were recorded for these solutions. Hydration number around Br^- was obtained from these spectra by means of the curve-fitting method. The results showed that the hydration number around Br^-was different with the concentration of the solution, and it changed when other cations were added to the solution. The same results were obtained from the molecular dynamics (MD) simulation, and the hydration number around Br^- can be reasonably explained by taking long-range structure into account.

View full abstract

An ultrasonic CT method is researched for measuring indirectly the interior temperature distribution from a series of sound projections which are taken at several different orientations relative to a phantom body. First, we compared three popular reconstruction algorithms to minimize a required number of projections. Then, phantoms were placed in a water bath, and a series of sound projections was measured when a pair of ultrasonic detectors were both translated and rotated. We could obtain the temperature distribution in phantom by subtracting the sound profile when non-heating from one when heating.

View full abstract

To investigate the pressure change and flame characteristics in a stretched, rotating flow, tubular flames of a lean hydrogen, methane, or propane-air mixture have been numerically simulated. Results show that, with rotation, the flame temperature of hydrogen and methane mixtures increases monotonically, while, that of a propane mixture decreases. With an increase of the fuel concentration, the position at which the reaction rate is maximum increases, and the temperature change becomes small. As seen in the pressure distribution, the pressure decreases around the center, and a pressure gradient is formed. This pressure gradient is steep near the center, but decreases as the radial distance is increased. The fuel flux decreases with the increase of circumferential velocities because of the decrease in the pressure gradient. For these reasons, this temperature change could be explained in terms of the pressure diffusion which results in mass transport due to the pressure gradient. However, it is also found that, with rotation, the pressure decreases and the density changes. The velocities increase due to flow expansion, resulting in an increase of flame stretch. Thus, the flame characteristic change with rotation is explained with the coupling of pressure diffusion and stretch effects.

View full abstract

A numerical simulation of a turbulent premixed flame stabilized by a pilot flame and bluff body was performed using the LES turbulent model and a new turbulent combustion model. In the previous work, the authors evaluated a local quenching effect due to a flame stretch and proposed a new combustion model including turbulent effects. The model also includes the temperature effect of unburned gas and the pressure effect in a combustor. First, the flame propagation to the premixed gas from the pilot flame was estimated. Due to the strong velocity gradient through the boundary layer between the premixed gas and the pilot flame, local quenching occurred around the top of the boundary layer. Thus, the ignition point of the premixed gas was located below the boundary region, where the velocity gradient was decreased. Next, the stable premixed flame after the flame propagation was estimated. Due to the strong organized motion created by the bluff body and the boundary layer between the premixed gas and the pilot flame, we could see fluctuations of the premixed flame. Further more, the distributions of the time averaged temperature and the intensity of the temperature fluctuation were compared with the experiment data. The calculated results were in good agreement with the experimental ones.

View full abstract

A liquid fuel (kerosene) jet was ejected across high-temperature and a high-speed airstream to elucidate the disintegrating process of the spray. All tests were performed at an airstream velocity of 30-80m/s and a temperature of 300-900K. Spray shapes were obtained by taking scattered-light photographs with a long exposure time. Short exposure photographs were used to obtain detail of the spray near the injector. Although the effect of parameter, the momentum flux ratio of ejecting liquid to airstream, on the heigh of the spray outer line could be explained for the room temperature airstream, it could not be explained for the high-temperature as follows. Height of the spray outer line increases with the temperature of the airstream under the condition of the same momentum flux ratio and decreases with airstream velocity under the condition of the same momentum flux ratio. Height of the outer line hardly changed when the airstream temperature increased from 500K to 900K under the condition of the same ratio of the mass flow rate of airstream to ejected fuel.

View full abstract

A bubbling liquid fuel (kerosene) jet was ejected normal to a high-temperature and high-speed airstream to control the spray shape and terperature of the airstream under an airstream velocity of 80m/s and a constant ejecting pressure. Spray shapes were obtained by taking scattered-light photographs with a long exposure time, the flame photographs were taken in the downstream region of the flame holder and the temperature of the airstream was measured at the exit of the test section of the wind tunnel. The height of the outer line of the fuel spray and the temperature profile of the airstream hardly changed and the temperature of airstream decreased when the ratio of the mass flow ratio of the ejected liquid Q_k to the standard mass flow rate Q_k0 decreases under the condition of Q_k/Q_k0 from 100% (without gas) to 56%. However, with decreasing in Q_k under the condition of Q_k/Q_k0 less than about 50%, the temperature profile of the airstream changed.

View full abstract

The Compression-ignition processes of a propane/air mixture are said to change from the cool flame and to the hot flame reactions. In this study, a rapid compression machine and propane were used to provide the cool flame and the hot flame reactions. At a constant injection timing of a diesel spray, the equivalence ratio of the propane/air mixture was varied in the range of 0 to 0.4 which resulted in determining the ambient state at the time of injection, the state of cool flame or the hot flame. Ignition, combustion, the KL factor and flame temperature obtained by a two color method were analyzed on a diesel spray formed in the ambient state.

View full abstract

In this study, the planar laser-induced fluorescence (PLIF) technique was used to measure the mean size and size distribution of diesel spray particles. The fuel used was n-tridecane mixed with 1 wt% N, N, N', N'-tetramethylparaphenyenediamine (TMPD). The light source used to excite the TMPD in the fuel was a secondary harmonic of a ruby laser-light sheet. A highly magnified image of the fluorescence from TMPD was taken by a 35mm still camera with magnified optics, and the mean particle size and particle size distribution of the fuel spray were determined by processing the images of fuel particles printed on paper. First, the accuracy of this method was confirmed by comparison with results of Phase Doppler Anemometry for fuel spray of an air-assisted gasoline injector. Then, for the diesel spray, the effects of injection velocity, ambient pressure, geometric configuration of nozzle hole (i.e., nozzle hole diameter and nozzle hole L/D) and of measurement points on the fuel particle mean size and size distribution in a high-pressure vessel at atmospheric temperature were investigated. The results showed that the small size particles increase in number with increasing injection velocity. At higher injection velocity, seem to atomize more actively. With increasing ambient pressure, the mean particle size increases. A reduction in nozzle diameter resulted in no improvement of atomization in this study. Also, the mean particle size in the downstream region of the spray is larger than that in the upstream region of the spray.

View full abstract

The generation of unburned hydrocarbon emissions due to physical effects was studied using KIVA-II. The results showed that exhaust hydrocarbons increase due to poor mixing of fuel with air during the ignition delay period at the low swirl case, and bulk quenching due to the high mixing rate and high heat loss to the wall during the combustion and expansion processes in the high swirl case. The re-entrant combustion chamber is advantageous for unburned hydrocarbon emissions in the case of retarded injection timing, but this modification produces excessive mixing and bulk quenching for advanced timing and, as a consequence, higher hydrocarbon emissions.

View full abstract

The effect of port water-injection on the exhaust NO_x reduction was examined experimentally as well as theoretically. In the experiment, water was injected into each suction port of a 4 cylinder turbocharged DI diesel engine using gasoline injectors. The exhaust NO_x was reduced significantly by port water-injection, and about 50% reduction in NO_x concentration was obtained by injecting an amount of 0.035kg of water/kg of air under various engine operation conditions. By comparing between the experiment and the analysis based on the authors' two-zone model, it is shown that the NO_x reduction is mainly caused by a decrease in the temperature of the burned gas resulting from an increase in the specific heat due to humidification.

View full abstract

The main factors which prevent quantitative soot measurements using laser-induced incandescence (LII) are expected to be as follows : laser intensity reduction due to soot clouds, LII signal attenuation due to soot clouds lying between the observation points and a camera and uncertainty of a relation between the signals and soot concentrations. To investigate the effects of these factors and solve these problems, LII was applied to a premixed flat flame where soot concentrations were measured in advance by transmissive light extinction method combined with computed tomography. The following points are clarified. (1) The correction on the incident laser attenuation is not necessary by using sufficient laser power because the LII signal saturates over a critical laser intensity. (2) The soot concentration is obtained from the LII signal by the calibration. (3) The signal attenuation can be corrected by multi-layer measurement tequnique using Lambert-Beer's law. Consequently, a quantification has been realized with accuracy of ±10% in this work.

View full abstract