The conventional Galerkin finite element solution is mesh dependent, and its discretization for Poisson's equation can not satisfy the conservation law at a nodal level when unstructured linear meshes are used. This research tries to solve these problems by introducing a new concept of the virtual nodal domain(Vnd) for a linear hexahedral element, and distributing the source term to a nodal algebraic equation in proportion to the volume of the Vnd. The Vnd is evaluated using a second-order flux existing within a linear element. We proved that the total Vnd of the eight nodes equals to the volume of the element, which guarantees that our scheme is also elementally conservative. Numerical simulation of heat conduction with both Dirichlet and Neumann boundary conditions shows that the accuracy has been improved obviously comparing with the conventional Galerkin FEM for unstructured hexahedral meshes, especially for bad quality elements. Our scheme can be introduced into any commercial FEM code quite easily.

View full abstract

The surfactant effect on the three-dimensional motion of a single bubble rising in stationary liquid is recently reported by Tagawa et al. (2010) as a first report. In this paper we investigate the forces acting on the single bubble under various surface slip conditions. Using our three-dimensional measurement system, velocities, trajectories and shapes of a bubble of 2.0± 0.1mm diameter in 1-Pentanol solution are obtained. We observe the quasi-steady helical motions without shape oscillations. These motions enable us to apply the generalized-Kirchhoff equation proposed by Mougin et al. (2002) for the calculation of the drag and lift forces. The slip condition is quantified as the normalized drag coefficient C_D^* which is the experimental drag coefficient normalized by drag coefficients of free-slip and no-slip conditions; C_D^* = 0 for free-slip, C_D^* = 1 for no-slip condition. It is found that maximum magnitude of lift force F_L is in C_D^* = 0.12 where the minimum frequencies f and maximum amplitudes A of the bubble horizontal path are seen. We show that the non-monotonic behavior of the frequency f on the slip condition can be explained by the direction of the lift force. Remarkably, for the non-zero C_D^* in surfactant solution the direction of the lift force is almost constant and lower than that at C_D^* = 0. It implies that the double-threaded wake keeps almost same alignment in the reference frame due to the constant slip condition along the azimuthal direction.

View full abstract

The aim of this study is to investigate on the flow behavior in the wake behind a circular cylinder by induced jet using dielectric barrier discharge (DBD) plasma. The plasma was created using a wave form with bipolar voltage of 4 kV_pp and frequency of 10 kHz, and the plasma actuators were mounted on the circular cylinder. The forward and backward jets for the main flow direction were induced by electrode arrangements of the plasma actuators. The flow field was visualized using a high-speed video camera, and the streamwise and vertical velocities ware calculated by dynamics PIV method at Reynolds number of 1.0×10³. We discussed that the vortex motion around the cylinder, Reynolds stress profiles and the half width of wakes by the forward and backward jets. Also, the characteristic frequency of vortex shedding was confirmed from the continuous photographs.

View full abstract

Preconditioning method developed by the authors was coupled with Building-Cube Method(BCM) which can simulate flows around a complex-shaped body. In this study, this method is applied to the simulation of thermal convective flows around an arbitrary-shaped body. First, forced and natural convections around a cylinder are calculated and the results are compared with the experiments and the previous numerical results to validate the calculated results. Second, natural convection around a fin-shaped body is calculated and the availability of this method to practical problems involving a thermal convective flow is discussed.

View full abstract

Unsteady two-dimensional flows in a steam turbine partial admission stage are numerically investigated using a CFD method developed by the authors. First, the flows through 54 nozzle and 30 rotor channels with a simple flange are calculated and the effect of the flange width on the rotor performance is predicted. Numerical results indicate that the reduction of flange thickness basically improves the operating efficiency and reduces the fluctuation of fluid force. Next, flange geometry with a sloped side wall is employed as considering a real design configuration and the flow is calculated. Finally, we discuss how the geometry is effective to reduce the fluctuation of fluid force and to improve the operating efficiency.

View full abstract

We developed a new passive-type lamination mixer for high viscosity fluid with a low Reynolds number, based on the baker's transformation (BT). BT is the best transformation for mixing fluids of laminar flow. However, there was difficulty in mass-producing the BT structure especially for micro devices like MicroTAS, Lab-on-a-Chip and Micro-Reactors, because conventional BT mixers require three-dimensional (3D) piping structures. We have successfully developed the easy-to-massproduce BT mixer by changing that concept of 3D piping structures to 3D channel structures. The 3D channel structures are not easy to produce by photolithography unlike the conventional mixers, while they can be easily mass-produced by molding once their 3D molds are produced. In this report, we newly developed a miniature scale BT mixer to meet the needs for mixing high viscosity fluids in food processing, resin blending, etc. An experiment for mixing different colored hardening silicone elastomers was performed by using the prototype mixer made of aluminum alloy, and the good BT mixing result was obtained, with observing several cross sectional patterns. The numerical fluid analysis also gave similar results of the patterns to those observed in the experiment.

View full abstract

In this study, the liquid plane jet with the second-order chemical reaction (A×B→R) is investigated experimentally. The reactants are 1-naphthol (A) and diazotized sulfanilic acid (B), and the product is 4-(4'-sulphophenylazo)-1-naphthol (R), briefly monoazo dyestuff. The main flow contains the species B and the water solution of species A is issued into the main stream by the plane jet. The blue dyestuff (C: Acid Blue 9) is also added into the jet flow and the concentration of species C can be treated as the conserved scalar (which is independent of the above chemical reaction). The concentrations of species C and the product species R are measured simultaneously by the optical fiber probe based on the light absorption spectrometric method, and the concentrations of species A and B are obtained from the conserved scalar theory. The streamwise velocity is measured by the I-type hot-film anemometer. We attempt to measure the concentrations of reactive species and the streamwise velocity simultaneously by the new combined probe which consists of the optical fiber probe and the I-type hot-film probe. First, it is examined whether it is possible to use the combined probe for simultaneous measurement. And then, the concentrations of the reactive species and the streamwise velocity in the plane jet are measured by the combined probe. It is found that the mass flux of reactant species involved in the jet flow (<uγ_A>) becomes larger and the mass flux of reactant species involved in the main flow (<uγ_B>) becomes smaller than nonreactive case due to the chemical reaction near the jet exit, while the influence of the chemical reaction on these mass fluxes in the downstream and outer regions is opposite to the influence near the jet exit. The results also show that the mass flux of product species (<uγ_R>) has negative value near the jet exit and positive value in the other region.

View full abstract

Joint statistics between streamwise velocity and concentration of reactive species in a planar liquid jet with a secondorder chemical reaction (A×B→R) are experimentally investigated. A water solution of species A is issued into a main stream containing species B. In this study, velocity and concentration are simultaneously measured by combining I-type hot-film anemometry with an optical fiber probe based on light absorption spectrometry, and we investigate correlation coefficients, joint probability density functions, and cospectra between velocity and concentration. The results show that the chemical reaction makes the absolute value of a correlation coefficient between u and γ_B small whereas a correlation coefficient between u and γ_A becomes large on the jet centerline owing to the chemical reaction, where u is the streamwise velocity fluctuation and γ_i (i = A, B, R) is the concentration fluctuation of species i. The correlation coefficient between u and γ_R has a negative value in the upstream region and has a positive value in the downstream region and the outer edge of the flow. It is also shown that the influence of the chemical reaction on the cospectrum of u and γ_i in the upstream region and near the jet centerline is different from that in the downstream region and the outer edge of the flow.

View full abstract

Investigations on the definition and randomness of turbulence was reviewed at first. Then, the Kolmogorov complexity which measures the randomness were introduced. Numerical and graphic data in the mixing layer which was formed downstream of two-dimensional nozzle exit were compressed with the aid of a compression program. Approximated Kolmogorov complexity, AK, and normalized compression distance, NCD, were obtained. The AK indicated the regularity of the laminar flow and the randomness of the turbulent flow quantitatively. The NCD of the numerical value varied with data length. Between the same data, it approached zero, yet, on the other hand, between different data, it approached unity as the data length increased. The NCD of the numerical value in the natural transition process in the mixing layer increased monotonically downstream. Thus the NCD appears to be the measure of the transition process. In the natural transition process in the mixing layer, the AK of the numerical value and the NCD of the graphic data did not change monotonously in the downstream direction. Thus they contain some uncertainty for the measure of the transition process.

View full abstract

The cavitation erosion in liquid metals is important for the safety and reliability of nuclear reactors and industrial machinery. However, an evaluation method of cavitation erosion in liquid metal has so far not been established. This study clarifies the incipient condition of cavitation in a liquid metal. A cavitating liquid jet test apparatus in a liquid metal was developed. A cavitating liquid jet was issued from an orifice nozzle with a hole of 0.6mm diameter. PbBi-68 was used as liquid metal. The tests were carried out in water using various nozzles with different shapes to obtain the nozzle shape effect on the incipient number and the erosion rate. It was found that the incipient cavitation number in tap water and PbBi-68 is almost the same regardless of the flow velocity. When the cavitation number is defined by an equation as 2(p_d-p_v)/ρv² (p_d: downstream pressure, p_v: vapor pressure, ρ: density, v: throat velocity), the incipient cavitation number becomes a constant regardless of the nozzle shape. Cavitation erosion tests were carried out with different nozzle shapes. It was concluded that the erosion rate can be qualitatively estimated by the output voltage of an acceleration sensor.

View full abstract

The calibration uncertainty of flowmeters using the subtraction method for hydrocarbon flow has been experimentally and analytically evaluated in order to calibrate flowmeters at lower flow rates than the flow rate range calibrated using the calibration facility. An expanded uncertainty is estimated to be 0.79 % at the flow rate ratio of 1/100 in the kerosene test rig of the primary standard facility. The main sources of combined uncertainty due to the subtraction method are the reproducibility of the standard flowmeters and the temperature measurements. Furthermore, a flowmeter has been calibrated using the subtraction method in order to verify uncertainty analysis. The result shows that the deviation of the calibrated flowmeter from the value using the primary standard agrees within the quoted expanded uncertainty.

View full abstract

Single-blade centrifugal impellers widely used as sewage pumps are geometrically asymmetric. Therefore, circumferential static pressure distribution of an impeller changes considerably with its rotation, and is considered to drastically change radial and axial thrust. Therefore, to improve pump reliability, it is necessary to quantitatively understand unsteady hydrodynamic force and elucidate the behavior and mechanism of the thrust generated. This study investigates the radial and axial thrust acting on a single-blade centrifugal impeller by conducting experiments and CFD analysis. As a result, the behavior of the radial and axial thrust of the hydrodynamic force acting on rear shroud and front shroud surfaces was clarified. The axial thrust of the hydraulic part is the lowest, and those of the rear shroud and front shroud are dominant compared with their radial thrust. Increase in the axial thrust of the shrouds is due to the increase in the leakage flow rate with the increase in the circumferential static pressure distribution of the impeller.

View full abstract

An analytical method was proposed and validated for droplet diameters and size distributions. The method was developed based on the energy conservation law including surface free energy and Laplace pressure. Under several hypotheses, the law derived an equation indicating that atomization resulted from a kinetic energy loss. Thus, once the amount of loss was obtained, the droplet diameter was able to be calculated without any experimental parameters. When the effects of ambient gas were ignorable, injection velocity profiles of liquid jets were the essential factor for the reduction of kinetic energy. The minimum Sauter mean diameter produced by liquid sheet atomization was inversely proportional to the injection Weber number under the conditions of injection velocity profiles with laminar or turbulent. By applying the mean diameter model, a non dimensional distribution function was also derived assuming Nukiyama-Tanasawa's function. The validity of these estimation methods were favorably confirmed by comparisons with corresponding mean diameters and the size distributions, which were experimentally measured under atmospheric pressure.

View full abstract

Because of the fast frequency response of the constant-temperature anemometer (CTA), it is useful for the measurement of the velocity fluctuation in turbulent flows. The paper proposes a method for attaining faster frequency response of the CTA to turbulent velocity fluctuations. Experimental calibrations show that the rolloff frequency of CTA can be improved from 5 kHz to 20 kHz by the present method under the condition that the diamter of the hot-wire is 5 μm and the velocity of the flow is 20 m/s.

View full abstract

The novel imaging technique combining Glare-Points technique (GPT) with interferometric laser imaging for droplet sizing (ILIDS) has been developed to measure droplet size, velocity and spatial distribution of mass flux in a spray. This technique enables to advance the droplet positioning error by ILIDS and expand the dynamic range of droplet sizing, because the large and small sized droplets were obtained by using GPT and ILIDS, respectively. The novel receiving optical system called as Double-planes Particle Imaging (DPI) was constructed in order to capture the focus plane (Glare-Points) and the defocus plane (interferometric fringe) with only one imaging sensor. Firstly, the captured image is processed for calculating of the droplet position, size and velocity by GPT and ILIDS, respectively. The droplet position by ILIDS is corrected by the mapping function which is obtained from the ray tracing simulation. Secondly, the same droplet image is matched with the aid of velocity information. In addition, the droplet position is defined as the one of GPT, the droplet size is employed the ILIDS' for smaller and the GPT's for larger than threshold value. Finally, the depthwise positions of droplet are estimated from the optical path difference between the focus and defocus images. From the result of validation experiment using a mono-disperse generator, the DPI enables to estimate the three-dimensional droplet position within the measurement error of 8.4% and expand the dynamic range of the accurate droplet sizing up to approximately 200 μm. Furthermore, this developed technique was applied to a spray flow, and the spatial distribution of mass flux was evaluated from the obtained droplet position, size and velocity information.

View full abstract

Oscillation behavior and heat transport characteristics in a pulsating heat pipe consisting of a single and straight tube with an open end are investigated experimentally. Periodic oscillation of a vapor plug with a large stroke over two thirds of the heat transport tube length is excited for heating powers larger than 20 W at horizontal orientation and continues for a long period over 8,000 s without cease. Heat transport rate and the effective thermal conductivity increase with heating power up to approximately 75 W and 40 kW/(m·K), respectively. Heating section, which has an inner diameter little larger than that of the heat transport section, is maintained at almost the saturation temperature of the working liquid throughout the succession of oscillation. Liquid film beneath the vapor plug is partially pushed back to the heating section to spread on the wick each time the vapor plug shrinks. The small step at the tube connection between the heating section and the heat transport tube section and the wick inside the heating section are shown to be essential structures for realizing the stable oscillation leading to high heat transport characteristics.

View full abstract

In the view of the safety operation and the designing of boiling two-phase flow equipments, boiling two-phase flow characteristics, especially the critical heat flux is very important. In addition, the flow obstacle in a boiling channel, such as the spacer of boiling water reactor, influences the boiling heat transfer and flow characteristics. In this study, the experimental investigation was conducted by using a SUS304 tube (I.D. = 8 mm, L_T =810, 840 and 900 mm) in which a rod-type flow obstacle (O.D. = 3.6 mm L = 20 mm) was equipped. Two kinds of detecting points of CHF, i.e. at the upstream of the flow obstacle and at the exit of the test section were observed. The difference of the detecting location of the CHF has been explained well by using the film flow model. On the basis of these results the influence length of the turbulent effect by the flow obstacle and the magnitude of the entrainment caused by the boiling were estimated.

View full abstract

In our previous research, a vibration mill using cog-disk mediums, which replaces the ball medium with a cog-disk medium in conventional vibration mill, was developed to achieve high-impact pulverizing in order to attain effective pretreatment of lignocellulosic biomass for producing the bio-ethanol. In this study, a vibration mill using cog-ring mediums, which improves pulverization efficiency and mill volume, was developed in order to make suitable woody powder for enzymatic saccharification. To clarify pulverization efficiency, shapes and size in grinding medium was addressed in pulverization of Japanese cedar wood. And also enzymatic hydrolysis of pulverization powder was carried out. As the results, cog-ring mediums showed effective pulverization at 800g batch milling. The average particle diameter was reduced from 55μm to 20μm in 20mintues. And the crystallinity index was reduced form 44% to 13% in 30minitues. Then saccaharification efficiency of holocellulose was reached over 70% at pulverizing time of 60min. This pulverization expanded the amount pulverizing volume forth to make high saccharification capability powder comparing with the existing vibration mill using cog-disk mediums.

View full abstract

To increase the electric power density of a polymer electrolyte fuel cell (PEFC), it is necessary to increase current density flowing through the membrane electrode assembly (MEA) in the whole PEFC. And this is controlled by the conditions of the fuel gas such as flow rate and humidity. We measured the spatial distribution and time dependence of changes in water content in the polymer electrolyte membrane (PEM) using Nuclear-Magnetic-Resonance (NMR) sensors in order to determine the influence of humidity and the utilization ratio of fuel gas on water content in PEM. Our experimental results showed that the water content at the outlet of the fuel gas increased with time when the flow rate was low, and this increase was induced by the water produced at the PEM. On the other hand, the water content in the PEM decreased with time when the flow rate was high and the humidity of the fuel gas was 70 %RH. When the humidity of the fuel gas was 85 %RH, we observed that water condensed at the outlet of the fuel gas. And a map illustrating the behavior of water content in PEM was made from the results.

View full abstract

To increase the electric power density of a polymer electrolyte fuel cell (PEFC), it is necessary to increase current density flowing through the membrane electrode assembly (MEA) in the whole PEFC. And this is controlled by the conditions of fuel gas such as flow rate and humidity. We measured the spatial distribution and time dependent changes of current density in the PEFC and the water content in PEM using Nuclear-Magnetic-Resonance (NMR) sensors in order to determine the effect of humidity and the utilization ratio of fuel gas. Our experimental results showed that the current density was uniformly distributed from the inlet to the outlet of the fuel gas when the flow rate was high and the humidity of the fuel gas was 70 %RH. On the other hand, when the flow rate was low and the humidity of the fuel gas higher (85 %RH), the spatial distribution of current density and water content in the PEM was non-uniform. We determined the relationship between the transfer phenomenon of water in the PEFC and this non-uniformity of water content and current density. In addition, a map illustrating the behavior of water content in the PEM and current density under electric-generating condition of the PEFC was made showing the relationship between relative humidity and utilization ratio of fuel gas supplied to the PEFC.

View full abstract

Dual fuel operation using different reactive fuels have been examined. The present study conducted experiments in a single cylinder compression ignition engine. N-heptane (octane No.=0) and i-octane (octane No.=100) were injected through two different injectors set at intake port and cylinder head, respectively. By comparing a case (A) in which n-heptane was directly injected into premixed charge of i-octane with a case (B) in which n-heptane was premixed and i-octane was injected into the n-heptane mixture, and by widely changing injection quantity ratio and direct injection timing, the effects of mixture concentration and combustion phasing on emissions and engine performance including thermal efficiency and maximum pressure rise rate were investigated.

View full abstract