Mass transfer characteristics of gas diffusion layer (GDL) are closely related to cell performance in polymer electrolyte fuel cell (PEFC). It is important to illustrate the water behavior in MPL for improving cell performance in operation. Because MPL is adjacent to catalyst layer and GDL substrate, supposedly influences their liquid water behavior. The paper with MPL was evaluated as specimen GDL. The images of liquid water distribution in the GDL were taken by neutron radiography under a few experimental conditions in order to elucidate the characteristics of water behavior in MPL. The images reveal that neutron radiography can be used to analyze the water behavior in MPL as an effective tool. It is clarified; MPL has two aspects, that is, one aspect is that the water retentivity and nature of network between MPL and paper substrate are formed under partially-saturated condition, another is that MPL acts as hydrophobic under dry condition.

View full abstract

In the first report, we had investigated the characteristics such as the changes of the mean velocity component profiles, half-widths (b_<1/2>) of the mean velocity profile in the downstream direction of the Push-Pull flow. In this report, we have experimentally investigated the characteristics of the turbulence intensity, Reynolds shear stress, and intermittency of turbulence. Turbulence intensity √<<u'>^^^-^2>/U_1 in the axial direction is not affected by the flow rate ratio Q_3/Q_1 (Q_1 is jet flow rate from mozzle and Q_3 is suction flow rate produced by the hood.). At the same time, turbulence intensity √<<v'>^^^-^2>/U_1 in the lateral direction becomes greater as the hood is approached and the flow rate becomes larger. These results are in accord with the values of production term for <u'^2>^^^- and <v'^2>^^^-. Although the variation of √<<u'>^^^-^2>/u^^-_m with respect to dimensionless coordinate y_1/b_<1/2> becomes similar with that for completely developed two-dimensional jet, the value decreases in the downstream direction as the flow rate ratio becomes larger for Push-Pull flow. But the flow rate ratio do not affect √<<v'>^^^-^2/U_1. Reynolds shear stress becomes smaller as the flow rate ratio becomes larger near the hood. Dimensionless distance y_1/b_<1/2>, from the centre axis of the flow to the point where intermittency factory γ becomes constant value (γ=0.3 or 0.5), becomes narrower as the flow rate ratio becomes larger near the hood. At the same time, the distance becomes smaller than that for completely developed two-dimensional jet.

View full abstract

The aim of this study is to investigate on frequency shift of vortex shedding from a circular cylinder by induced jet using dielectric barrier discharge (DBD) plasma. The electrode of a plasma actuator was mounted on the circular cylinder. A wave form with voltage of 2kV and frequency of 10kHz is applied to the electrode. The forward and backward jets for the main flow direction were induced by electrode arrangements. The time-averaged velocity, turbulent intensity, half width and the vortex shedding frequency in the wake behind the circular cylinder were measured using a I-type Hot-wire anemometer and flow field is visualized using smoke wire method and a digital video camera at a Reynolds number of 1.0×10^3. We discussed that the turbulent intensity of backward jet increases near the circular cylinder, and the half width of wakes decreases and increases using the forward and backward jets. Also, the frequency shift of vortex shedding from circular cylinder using plasma actuators changes with respect to no jet.

View full abstract

A Cartesian-mesh approach named Building-Cube Method (BCM) was developed for flow computations. This method has a feature of dividing a computational domain into a number of sub-domains (named as 'Cube') by Quad-tree or Oct-tree, and processing the sub-domains in parallel. In this paper, we propose the modified BCM for changing the subdividing method of the computational domain from Quad-tree to Nona-tree (in two-dimensional case). The purpose of the modification is to subdivide the computational domain more efficient and also to simplify the data exchange method between cubes. The method is evaluated by computing flows around a circular cylinder and an airfoil. It is shown that the modified BCM divides the computational domain more efficiently than the original BCM without degrading the computational accuracy. It is also concluded that the mesh size of about 0.3 Re^<-0.5> near the surface is necessary to reduce the influence of the staircase approximation of surface in the Cartesian-mesh approach.

View full abstract

Turbulence characteristics in the spatially developing fractal-generated turbulence are investigated by means of the direct numerical simulation (DNS). The square type fractal grids are numerically constructed using the immersed boundary method. The effects of solidity σ and thickness ratio t_r, which are ones of the most important grid parameters for fractal grids, are investigated in detail. The results show that the increase of σ leads to the improvement of the uniformity and isotropy of flow fields. The integral length scale in the decaying region of the fractal-generated turbulence remains monotone increase at high σ and t_r, whereas it is approximately constant at low σ and t_r. The derivative skewness and flatness in the streamwise direction are different from those in the transverse direction for all σ and t_r investigated. These results will be useful for designing fractal grids for practical applications.

View full abstract

For the achievement of low-cost computations of the fluid flow through micro-nano sized porous media such as catalyst layer of fuel cell, lattice Boltzmann method (LBM) has been improved for high Knudsen number flows by considering the diffuse scattering boundary condition, effective relaxation time related to Knudsen number and the regularization procedure. These treatments have been validated in simple geometries such as two-dimensional Couette flow and channel flow. In this paper, above micro-flow LBM is extended to three dimensional flow. A nano-mesh structure is presumed and its applicability is discussed in terms of discrete velocity (Q19 and Q39 models) by comparing to the molecular dynamics simulation. It is found that the resulted velocity profile by LBM agrees with that by molecular dynamics including the velocity slip at the porous wall. The Q19 model has enough accuracy compared to Q39 model, which suggests that the advantage of the simple model for the application. The less difference between Q19 and Q39 models implies that the third order term in Hermite polynomials becomes less effective for the complicated flow regime.

View full abstract

The unsteady behavior of a vortex generated on the diffuser leading-edge, which is called the leading-edge vortex (LEV), is discussed through experiments and numerical simulation. The LEV is different from the separating vortex of the diffuser leading-edge and passage vortex of the diffuser, develops rapidly with a decrease in the compressor mass flow rate, and forms a flow blockage in diffuser passages. Therefore, the evolution of the LEV may become a cause of diffuser stall. Additionally, in one attempt to control the LEV, two types of tapered diffuser vanes, which are shroud- or hub-side tapered diffuser vanes, were adopted. Though the shroud-side tapered diffuser vane can effectively reduce the compressor noise level, the compressor performance deteriorates remarkably. On the other hand, the hub-side tapered diffuser vane not only reduces the compressor noise level but also improves the compressor performance. According to the visualization results of the oil-film methods and numerical simulations, the hub-side tapered diffuser vane can suppress the evolution of the LEV in the compressor low-flow-rate operation.

View full abstract

A theoretical approach to the rheological properties of a particle model which is widely applied as a numerical method to analyze the behaviors of a continuous media has been constructed. The present study treats a particle system which describes the mechanical behaviors of the continuous media by means of the computational mechanical simulation. The behaviors of the particles in the system have been investigated analytically in order to conduct the numerical analysis effectively. A basic equation which describes the motion of the particle is derived from the particle system. Since the basic equation is the nonlinear equation which contains the thermal effect, it is analyzed theoretically by using some mathematical techniques. As a result, the time dependent behaviors of the stress and strain which are derived from the particle system are shown analytically. Especially, it is found that the viscoelastic behaviors emerge from the interactive behaviors of the particles under the constant temperature.

View full abstract

Numerical analyses are performed for attenuation of weak shock waves generated by explosion of a unit mass of TNT. Viscous and heat conduction loss and molecular vibrational relaxation effects of oxygen and nitrogen molecules are considered. The effect of each diffusion term on the wave attenuation and the wave form change is investigated in detail. In addition, the characteristics of each attenuation effect are investigated by the technique of frequency analysis.

View full abstract

A three-dimensional direct numerical simulation (DNS) with Level-Set method is applied to the impingement process of a single droplet with diameter of 1.2〜2.2mm on the air-water free surface, and the characteristics of vortices induced below the free surface are investigated by comparing with the experiments. The results show that the present DNS can capture the generation and transportation process of the vortices. The predicted vortex intensity corresponds with the experimental bestfit curve proposed for a larger droplet with diameter of 2.2〜5.6mm. In addition, the comparisons of velocity vectors and scalar concentration fields between the vertical and oblique impingements show that the momentum and mass transfer mechanism are strongly affected by the angle of droplet's impingement.

View full abstract

Three-dimensional direct numerical simulation (DNS) is applied to a gas-liquid two-phase flow with a non-mobile wavy gas-liquid interface, and the turbulence structure in boundary layers on both gas and liquid sides and the scalar transfer mechanism across the gas-liquid interface are investigated. The results show that the scalar transfer across the non-mobile wavy gas-liquid interface is mainly controlled by the longitudinal vortices related to bursting motions on the liquid side, whose mechanism is similar to those across the flat and realistic wind-driven gas-liquid interfaces. However, since the scalar transfer is strongly affected by the fixed interface configuration, the low-scalar-flux streaks observed in the cases of the flat and realistic wind-driven gas-liquid interfaces do not clearly appear. This means that in order to precisely predict the scalar transfer rate across the realistic wind-driven gas-liquid interface, the non-mobile wavy gas-liquid interface is not suitable to be employed.

View full abstract

A variable displacement pump is a high-efficient pump for hydraulic power steering systems. In order to develop a quiet variable displacement vane pump which leads to lower vehicle interior noise, we need to reduce pressure pulsation and vibration of the pump. So, we have developed a new simulation method for predicting pressure pulsation and vibration caused by a variable displacement pump. This simulation method combines hydraulic simulation and vibration simulation to predict vibration in the design phase. To improve accuracy of the hydraulic simulation, the deformation of the pump shaft was modeled. Moreover, method of characteristics using block diagram was used to calculate pressure pulsation of rubber-made pump outlet hose. This hydraulic simulation results were used as excitation force acting on the pump shaft in vibration simulation. As a result, hydraulic simulation predicted pressure pulsation of 11th order which comes from the number of vanes and its harmonics of the pump revolution accurately. The differences of predicted and measured results were within 0.1dB at 1500min^<-1>, 9MPa. And vibration simulation using the results of hydraulic simulation predicted vibration of 11th and 22nd order of the pump revolution that agreed well with experimental results.

View full abstract

The present study is concerned with numerical simulation of the movement of gas progressing with the constant velocity in a horizontal channel filled up with water. The dependence of the gas-liquid interface shape and the liquid flow in the neighborhood of the gas phase tip on the channel height and the gravity force are analyzed by the numerical simulation program FIDAP. The three cases that the height of the channel is greater than, almost the same as and less than the bubble limiting thickness are studied. The relation between the residual liquid film thickness on the wall after passage of the gas phase tip and the tip velocity is also discussed.

View full abstract

To estimate an influence of velocity on thermodynamic effect, we conducted experiments, in which inducer rotational speed was changed in liquid nitrogen. The experiments in liquid nitrogen and in cold water allowed us to estimate the amplitude of thermodynamic effect. In the experiment with lower rotational speed, the suction performance was improved in liquid nitrogen. The cavity length at lower rotational speed was shorter than that at higher one at the same cavitation number. Thus, we have confirmed that the degree of thermodynamic effect depends on the rotational speed as a suppression of cavity length. The temperature depression of liquid nitrogen was estimated from the comparison of cavity length between in liquid nitrogen and in cold water. We found that the amplitude of temperature depression became smaller when the rotation speed was lower. In addition, from arrange data using modified non-dimensional thermodynamic parameter Σ^*_<mod>, the effect of nonlinearity of vapor pressure/temperature curve should be considered to evaluate the thermo-dynamic effect especially in the case of higher temperature and rotational speed.

View full abstract

The self-consistent equations of the piston effect are used to calculate the behaviors of the unsteady airflows caused by the moving trains in the subway system. The system is composed of the tunnel for trains, pathways or stairways for passengers, and ducts for the ventilation or the air-conditioning, and strongly has the one-dimensional aspects. In order to make the one-dimensional pipe line net-work, every component of the system is transposed to the equivalent pipe, and is connected up each other to form the three station model. By making use of the theory of the piston effect and the interfering model of airflows with division or combining, the dynamic behaviors of the airflows are calculated as the computer simulation, and are compared with the field data measured in the real subway system. The result of simulation has a good agreement with the field measurements.

View full abstract

It is particularly important to know the deformation behavior of the erythrocyte passing through a micro channel (MC) like the blood capillary. In this paper, the deformation behavior of an erythrocyte passing through the pseudo MC was calculated and observed by a two-dimensional numerical simulation of blood flow. The pseudo MC wall was created by assuming that the wall was one of the fluids which had no velocity at all times. The erythrocyte's movement and deformation were calculated by using an immersed boundary method. The following results were obtained by the simulation. An erythrocyte smoothly deformed and passed through the pseudo MC. In the pseudo MC, its shape changed to bullet-like shape and its moving velocity was almost constant. In addition, a narrow layer of only blood plasma, what we call the plasma layer, was formed between pseudo MC's wall surface and the erythrocyte membrane. It was possible that the plasma layer puts the erythrocyte into fluid lubrication. On the other hand, in our previous researches the deformation behavior of the erythrocyte passing through the MC was observed by experiments. The simulation results had good agreement particularly on an erythrocyte's deformed shape and process in MC with the experimental results.

View full abstract

We developed a computational model that expresses a deformation behavior of a red blood cell (RBC) in a high shear flow. The RBC was modeled as a capsule of spring-networks that has a mechanical nature of lipid bilayer and spectrin. Besides bending and stretching of the membrane, area and volume constraints were imposed in order to assure their incompressibility. Fluid forces exerted by external plasma and internal hemoglobin were estimated based on the momentum conservation and Newton's law of viscosity. Given a flow field, the behavior of RBC was determined toward the minimum energy state. In the Couette flow, the RBC was stretched with a tank-tread rotation of the membrane. A good agreement with experimental data was obtained by using a nonlinear spring for the stretch where a spring constant changes as a function of stretch ratio. In the cyclically reversing unsteady shear flow at a frequency of 1, 2, 3 and 5Hz, the RBC deformed in concert with a change in the flow field. The time lag on the RBC deformation as well as the deformation index well agreed with experimental results both qualitatively and quantitatively. Those results demonstrate capability of the proposed RBC model.

View full abstract

The carotid bifurcation is found to be a major site of Atherosclerotic plaque formation and intima-media thickening. In the present study, oxygen mass transfer in the human carotid bifurcation has been numerically investigated, focusing on effects of the bifurcation angle and the volumetric flow ratio between the internal carotid artery (ICA) and the external carotid artery (ECA) on the magnitudes and distributions of the oxygen wall flux. Three-dimensional models of bifurcations with two different bifurcation angles mimicking the real bifurcation anatomy were constructed to perform simulations of steady blood flow under the wall boundary condition of a constant oxygen concentration. Results reveal that the axial flow separation at the outer common-internal carotid wall can significantly alter the distribution and the magnitude of the oxygen wall flux, depending strongly on the bifurcation angle. The magnitude of the lowest Sherwood (Sh) number (non-dimensional oxygen wall flux) at the ICA sinus, where Atherosclerotic plaque is likely to develop, is sensitive to the change in the value of the volumetric flow ratio rather than the change in the bifurcation anatomy.

View full abstract

A new method to actuate a small droplet on a plate was proposed. In order to reduce the resistance to the droplet movement, which is typically represented by the resultant of surface tension acting on the three-phase contact line, i.e., σ (cos θ_R-cos θ_A) (σ: liquid surface tension, θ_R, θ_A: receding and advancing contact angles), a supersonic oscillation was imposed on the plate to suppress the contact angle hysteresis (θ_R-θ_A). The experimental results using SAMs (Self-Assembled Monolayers) plates showed that the resistance was reduced by 80% due to the oscillation having the frequency of 28kHz and the amplitude of few micrometers. Then the Laser beam was directed to one end of the droplet. The contact angle was remarkably decreased due to the local heating of SAMs plate and the droplet moved to the end of the Laser beam, since the tangential component of surface tension at the heated end overcomes that on the other. The droplet can be actuated by about 0.6mm/s by the method combining the oscillation and the local heating by the Laser beam.

View full abstract

Ultrasound Velocity Profiler (UVP) is applied to measurement of two-dimensional velocity vector field of vortex rings in water. The instantaneous flow field passing through the measurement section is reconstructed from spatio-temporal distribution of velocity component obtained along two different ultrasound measurement lines. The method of reconstruction is valid when vortex rings have axisymmetric structure having constant translational velocity at the measurement section. Using this principle, vortex ring subject to background rotation is also presented to discuss the effect of weak Coriolis force on formation process of vortex rings.

View full abstract

We have conducted laboratory experiments on air-bubble generation by a submerged two-dimensional body moving at a constant velocity U_0 in order to understand the fundamental flow physics of marine constructions such as pipe lines close to the air-water interface. Measurements of the free surface profile and visualization of the air bubbles yield the threshold and the regime diagrams of the bubble generation, which are described by cross-sectional shape of the cylinders, Reynolds number (Re_d=U_0d/ν, where d and ν are the vertical thickness of the cylinders and kinematic viscosity of the water, respectively), Froude number (Fr_h=U_0/(gh)^<1/2>, where h is the depth of the cylinders), and normalized depth of the cylinder (a=h/d). For circular cylinders, as Re_d and Fr_h numbers increase, the surface deformation becomes substantial in the downstream of the cylinder and breaking wave with air entrainment occurs. The bubble generation by the breaking wave is also observed in the case of the elliptic cylinder although the condition of the bubble generation in the Re_d-Fr_h regime diagram is different from that observed in the case of the circular cylinder. In any cases of the cylinders, the ratio of the wave height to the wave length, which should be the physical criterion of the breaking wave with bubble generation, is about 0.1.

View full abstract

We investigated the interaction between microbubbles and Taylor-vortices, which are generated in a fluid layer between coaxial-rotating double cylinders. O (10μm)-diameter hydrogen bubbles were generated by water electrolysis and dispersed into the fluid layer. The maximum void fraction, which is estimated by the input power for the water electrolysis, was smaller than 0.02%. From the time averaged velocity distribution, which is observed by Ultrasonic Velocity Profiling (UVP), these values are changed with different tendency at Re/Re_c=3.0 or 4.0, where Re_c is the critical Reynolds number for onset of the primary instability. As initially expected, the existence of the microbubbles does not modify the axial wavelength of the vortices and the frequency of the azimuthal waves either. However, the power of the modulation wave component, which comes from the flow instability, is lowered by the addition of microbubbles at Re/Re_c=8.0. This phenomenon is caused by the rising bubbles with the pattern, which makes inhomogeneous local void fraction.

View full abstract

Annular two phase flow is encountered in many industrial equipments, including flow near nuclear fuel rods in boiling water reactor (BWR). Especially, disturbance waves play important roles in the pressure drop, the generation of entrainments, and the dryout of the liquid film. Therefore, it is important to clarify the behavior of disturbance waves and base film. However, most of the previous studies have been performed under atmospheric pressure conditions that provide the properties of liquid and gas which are significantly different from those of a BWR. Therefore, the effect of properties in gas and liquid on liquid film characteristics should be clarified. In this paper we focus on the effect of gas-liquid density ratio on liquid film thickness characteristics. The experiments have been conducted at four density ratio conditions (ρ_L/ρ_G=763, 451, 231, and 31). As a result, it was found that liquid film thickness characteristics including the effect of liquid/gas density ratios were well correlated with a gas Weber number and the liquid Reynolds number in the wide range of experimental conditions (ρ_L/ρ_G: 31〜763, We: 10〜1800, Re_L: 500〜2200).

View full abstract

A finite volume method (FVM) on distorted meshes for discretizing energy equation's conduction term is presented. In this method, it is possible to compose the computational mesh of general quadrilateral elements, namely, the cells are not necessarily to be rectangular. Gradient of temperature on the cell's surface is computed to be second-order accurate. Therefore, the error of numerical results by this method is smaller than using traditional multilateral element method (MEM). The error doesn't depend on the degree of mesh distortion. The formulation only based on Taylor's theorem is straightforward. These are advantageous features to revise the fluid flow computation programs (based on FVM) which neglected heat conduction term of energy equation. The test calculations show that the convergence tendency of the numerical error using this method on the distorted mesh is the same as using ordinary 2-node discretization on the constant-interval rectangular mesh. By this method conduction term was added to energy equation of SALE program which had neglected that term originally, and numerical calculation of a fluid flow with heat transfer problem was performed. The numerical result well agrees with the analytical solution.

View full abstract

This paper presents a numerical analysis method for solving shape optimization problems of domains in which steady-state heat conduction fields considering temperature dependency of thermal conductivity coefficient are defined. In this paper, we formulated two shape optimization problems, namely, maximization of thermal dissipation on heat transfer boundaries and minimization of the heat conduction fields. The shape gradient functions for these shape optimization problems were derived theoretically using the Lagrange multiplier method and the formulae of material derivative. Reshaping was accomplished using the traction method that was proposed as a solution to the shape optimization problems. The validity of the proposed method was confirmed by the results of 2D numerical analysis.

View full abstract

In either cases that liquid contacts with hot surface or it is rapidly superheated, homogeneous nucleation generation generally occurs in metastable liquid. The model proposed by authors elucidated the homogeneous conditions for both cases. A lower limit temperature under which the homogeneous nucleation never occur, is discussed on the basis of the model. As the result, the lower limit of the temperature on the interface is obtained to be T*_i^=303℃ for any water temperature from 0 to 100℃ at atmospheric pressure. For the case of ramped heating, the homogeneous nucleation generation always appears on the ideal surface without any cavity. In addition, the time at which the homogeneous nucleation generation takes place, strongly depends on the rate of temperature in crease. In additio, the average cluster temperature slightly varies from T_<avg>^*=302 to 313℃ according as the heating rate from 10 to 10^9K/s.

View full abstract

In order to spread economically viable distributed generation systems for apartment buildings, it is essential to develop an efficient and low-cost heat supply system. We are developing a new cogeneration system (Neighboring CoGeneration system: NCG system). The key concept of this system is to install a heat storage unit in each house for hot water supply and room heating and to connect heat storage units by a single-loop hot water pipe. The system leads to time leveling of the total heat supply and reduction of installation cost. Furthermore, it is expected that the cogeneration can operate according to electricity demand because of the large heat capacity of accumulation of the system. In this study, a dynamic simulation model was developed to evaluate the performance of the heat supply system and the environmental load-reduction effect of the NCG system for 50 households. It showed that the NCG system can supply enough heat for peak demand in winter and decrease 23% in CO_2 emission on yearly average.

View full abstract

To improve a performance of PEFC, it is necessary to maintain high current density generating in the whole area of a fuel cell. Spatial distribution of the electric current generating in PEFC can be measured from the frequency shift of Nuclear Magnetic Resonance (NMR) signal received by planar surface coils inserted into PEFC. Applying this technique, the frequency shift of NMR signal in the PEFC which sent power generation current in the lamination direction was measured. In this study, two kinds of Polymer Electrolyte Membrane (MEA) were used. One was MEA which applied the platinum catalyst to an area of 50mm×50mm. The other was MEA with platinum catalyst of its half area. The distributions of the frequency shift of NMR signal in PEFCs used two MEAs were measured, and the spatial distributions of current density generating in PEFCs were obtained. The result of MEA with Pt catalyst of 50mm×50mm area showed uniform current distribution in the whole area. On the other hand, the results of MEA with Pt catalyst of the half area showed two areas, the region of uniform current, and the region of zero current. These measurement results were in agreement with the theoretical-analysis results of magnetic field in PEFC. The uncertainly of this measurement technique was evaluated from the obtained results.

View full abstract

Gas diffusion layers (GDLs) coated with a hydrophobic microporous layer (MPL) have been commonly used to improve the water management property of polymer electrolyte fuel cells (PEFCs). In the present study, the influence of a hydrophilic and hydrophobic double MPL on PEFC performance under no-humidification at the cathode was evaluated. The hydrophobic MPL using a PTFE (polytetrafluoroethylene) binder and the hydrophilic MPL using a PVA (polyvinyl alcohol) binder were coated on the carbon paper GDL substrate. The thin hydrophilic layer coated on the conventional hydrophobic MPL enhances the ability to prevent drying-up of the MEA, thereby enhancing PEFC performance. The hydrophilic layer is effective for conserving membrane humidity. The hydrophobic intermediate layer between the hydrophilic layer and the GDL substrate prevents removing the water in the hydrophilic layer via dry air in the substrate. Reducing the hydrophilic layer thickness to 5μm is effective for enhancing PEFC performance. Appropriate enhancement of hydrophilicity by increasing the PVA content to 5mass% is also effective for enhancing PEFC performance.

View full abstract

The transient technique using infrared-thermography or liquid-crystal has widely been used for measuring the distribution of local heat transfer coefficients. In this technique, wall surface temperature is measured, and the heat transfer coefficient is calculated so as to accord the measured temperature with the theoretical solution of one-dimensional heat conduction problem. In actual cases of complicated surface geometry, however, three-dimensional heat conduction, caused by the three-dimensionality of the wall surface and the distribution of heat transfer coefficient, occurs in the wall. In this study, the heat transfer coefficient on the hemispherically dimpled surface was measured with an infrared camera, while the three-dimensional heat conduction in the wall was numerically calculated. In the compensation process, modification of the heat transfer coefficient was repeated till the numerical result agreed with the measured surface temperature. The present results showed that the heat transfer coefficient near the dimple edge was overrated, while that within the cavity was underrated. The maximum error induced by the three-dimensional heat conduction was 60% on the leading edge of the dimple, and the error in the other area was about 20% at most. At the dimple edge, the convex geometry increased the surface area where the heat flew into the wall, and consequently temperature rise became larger than the flat part. On the other hand, within the dimple, the concave geometry formed the radially-expanding heat conduction area, and the temperature became lower. The principal factor contributing to the error of the measurement is the three-dimensionality of the surface.

View full abstract

Jet flow from an orifice nozzle is applicable to mixing and entraining of ambient fluid effectively because of a large shearing stress layer with a large velocity gradient at the jet edge due to the vena contracta effect, but the sudden contraction at the nozzle exit accounts for a large flow resistance. The centerline velocity increases to the downstream and reaches to the maximum of 1.2 times of the nozzle exit maximum velocity at about x/d_0=2 (d_0: nozzle exit diameter), this is also effective to enhance the heat transfer performance of impinging jet. In order to reduce the flow resistance remaining a vena contract effect, use of a cone orifice nozzle has been considered. In this study, the effects of use of a notch for the orifice nozzle on the flow characteristics are examined experimentally. Hot wire measurements were conducted to demonstrate the spreading or mixing performance of the notched orifice nozzle having a reduced flow resistance of the nozzle and an increasing turbulent intensity more than that for a conventional orifice nozzle. Moreover, heat transfer performance of the impinging jet from an orifice nozzle is made clear and the improvement or enhancement by a notched orifice nozzle and a tapered notched nozzle will be shown by experimental research works.

View full abstract

Laser-induced ignition gathers attention as the relatively new technique of ignition owing to its flexibility of ignition point that could be of benefit for stability and combustion efficiency in the lean premixed combustion. In this study, the laser-induced ignition phenomena have been investigated experimentally, especially on the relation between the initial temperature of the premixed gas and the minimum ignition energy. Methane/air mixture is used as the premixed gas. The initial temperature of premixed gas is varied from 300K to 450K under the constant molar density. Results clearly show that the absorption energy to the plasma is not affected by the increase of the initial temperature of premixed gas. In addition, the minimum incident energy for ignition decreases with the increase of initial temperature of premixed gas. Eventually, the increase of the initial temperature improves the ignitability of premixed gas by the laser-induced ignition.

View full abstract

A scaling model for deflagrative blast decay is developed based on the Strhelow's constantvelocity flame model and the field experimental records of pressure transducers and high-speed video images of spherical flames in hydrogen-air clouds. The model provides a scaling rule in which the blast over-pressure normalized by the observed flame propagation velocity decays inversely proportional to the 0.73 power of the distance normalized by the characteristic explosion length. A similar decay behavior is also obtained for the normalized positive impulse. The impulse decay curves for lean and rich gas mixtures do not coincide, although the decay rates are close values.

View full abstract

The purpose of this study is to develop an ultra-micro-combustor that uses two types of coaxial cylindrical flames, rich premixed flame and diffusion flame. The combustor consists of inner and outer porous tubes, and rich propane-air mixture and air issued, respectively, through the inner tube outwardly and through the outer tube inwardly, forming a cylindrical stagnation plane sandwiched by the inner rich premixed flame and the outer diffusion flame. Petal type flame was also observed in the downstream of the cylindrical flames. Keeping the equivalence ratio φ_i and flow rate q_i of the rich mixture, air flow rate q_a was varied. The O_2 and CO concentrations and temperature of the burnt gas were measured, and heat loss ratio η_<hl> and combustion intensity L were evaluated. The obtained results are described as follows. (1) The relation curve of η_<hl> with the overall equivalence ratio φ_<all>, which is evaluated from the total flow rate of the fuel and the air, has a minimum value. (2) The relation curve of the minimum value of η_<hl> with L have a minimum value. (3) The CO concentration of the burnt gas increases as q_a is increased because of local extinction of the petal type flame. (4) When q_a is increased further, petal type flame is extinguished. After that, the O_2 concentration increases and the CO concentration decreases.

View full abstract

We made a study on the combustion chamber deposit (CCD) formation pattern on a piston crown surface in a small two-stroke spark ignition engine. Generation of the CCD is influenced by the temperature, pressure, gas flow conditions or retention period of the lubrication oil in the cylinder. Lubrication oil flows were observed by generating CCD streaks (CCDS) from many small drilled holes on the piston crown. To understand the temperature influence on the heat degradation of the lubricant oil, the oil on the conical cavity was heated statically in an electric furnace, and the oil was also dropped on the heated incline plate. Using separated a piston crown and a cylinder head, the characteristic properties of deposited sediments in the combustion chamber were surveyed and analyzed in a such operation load. In the results, (1) the oil flow-pattern of a piston surface flow is the average direction of gas flow there, (2) Most of CCD is formatted at about 275℃, (3) CCDS formed on the heated plate along the oil flow line is a distant every angle, (4) CCD on higher-heating area is formed heat-proof in a higher-boiling point-composition concentrated by a metallic detergent, and (5) CCD is many on lower-middle operation load and less on higher operation load.

View full abstract

The present study experimentally investigated the performance and emission characteristics of a diesel engine with hydrogen added to the intake air at late diesel-fuel injection timings. The diesel-fuel injection timing and the hydrogen fraction in the intake mixture were varied while the available heat produced by diesel fuel and hydrogen was kept constant at a certain value. As the hydrogen fraction was increased, NO first deceased, attained minimum, and then increased. The maximum rate of incylinder pressure rise also showed minimum at 10vol% hydrogen fraction. The indicated thermal efficiency was almost constant or slightly increased with small amount of hydrogen. A combination of hydrogen addition and late diesel-fuel injection timing contributed to low temperature combustion, in which NO decreased without the increase in unburned fuel.

View full abstract

As suggested by the equivalence ratio-temperature (φ-T) diagram, soot could be burnt off at 1800<T<2 200K where there is little formation of nitrogen oxides (NO_x). Accordingly, nonsooting and low NO_x combustion would be possible with moderate rates of exhaust gas recirculation (EGR), and without severe deterioration in combustion efficiency. However, complete burning-off of soot particles needs both sufficiently high temperatures and long residence times that would be favorable for NO_x formation. Therefore, a detailed examination of the kinetics of soot particle oxidation and NO_x formation processes would be useful. This study employs the detailed mechanisms from gas reactions to soot particle formation and oxidation, and a two-stage model is developed and used to investigate the trade-off between soot and nitrogen oxides emissions. The effects of the equivalence ratio, EGR, ambient pressure and temperature, and initial particle diameter are observed for various residence times. The results show that high rates of NO_x formation are unavoidable under conditions where high reduction rates of soot particles are obtained.

View full abstract

This Silent Discharge type of DPF (Diesel Particulate Filter) has been studied for eliminating PM (Participate Mater) we call it "SDeDPF". Usually, exhaust gas temperature of diesel engines is under 200 or 250℃ at normal city driving condition. Under that condition, generally PM is not bourn out in the normal ceramic DPF. This SDeDPF aims to remove PM electrically and chemically even at room temperature and atmospheric pressure continuously. Finally, in the basic lab test result, 95.6% reduction of PM has been verified by SDeDPF with a special MFS (Metal Fiber Sheet) for discharge electrode to reduce a back pressure, a special Turbulent Block for turbulent and slower velocity of exhaust gas, the 1mm gap between electrodes and an optimum total area of piled electrodes. Also, 98.1% reduction of PM could be designed by most suitable gap between electrodes.

View full abstract