Flow fields over annular plug nozzles are computationally simulated and the boat-tail drag characteristics are clarified and discussed based on the simulation results. The plug nozzle configuration is taken from the ATREX (Air Turbo Ramjet) engine under development at the ISAS. The simulations are carried out for the axi-symmetric nozzle configuration using compressible Navier-Stokes equations. The computed result shows that there exists a low pressure region downstream of the Prandtl-Meyer expansion on the boat-tail region and the separation shock wave appears due to the interaction of the main flow and the nozzle plume downstream. The result also shows that the primary cause of the boat-tail drag turns out to be the low pressure region downstream of the Prandtl-Meyer expansion. The flow field with the secondary flow injection is then simulated and the computed result shows that the pressure in the region in front of the separation shock wave increases and the secondary flow injection successfully reduces the boat-tail drag. The effect of the turbulence model used in the computation is discussed to ensure the reliability of the solutions.

View full abstract

The fourth order central finite difference scheme is applied to the convection terms in three-dimensional Navier-Stokes equations to simulate directly the flow fields in a medium Reynolds number baffled stirred tank. The rotational cylindrical coordinates with angular velocities are used to express boundary conditions on impellers. On the other hand the stationary cylindrical coordinates are adopted for baffles. When the fluid is water, the phenomena during 60 seconds, which are the sufficient time for obtaining various flow mechanisms and time averaged flow characteristics in a stirred tank, are simulated using this calculation method. The calculated results are in good agreement with the measured data. Then, this study makes a way for direct numerical simulations of various type stirred tank with various type impellers and baffles.

View full abstract

Two-dimensional steady flow of an incompressible viscous fluid around a circular cylinder in the case where the velocity field at large distances is the combination of a simple shear and a uniform stream is described in terms of matched asymptotic expansions valid at a low Reynolds number. The main purpose of the present paper is (1) to examine the validity of the assumptions used by Bretherton (1961) and (2) to construct an alternative approach without using such assumptions. In the present paper is constructed a system of governing integral equations for vorticity and stream function based on an Oseen-type equation. Local solutions, inner and outer solutions, are obtained from these equations by using the method proposed by Kida (1991), which is so systematic that we do not need the detailed physical consideration. Finally aerodynamic forces are compared with those obtained by Bretherton. The present paper shows that Bretherton's assumptions are correct within the first approximation. One cycle higher order solutions are obtained in this paper.

View full abstract

A finite element method for analysis of two-dimensional, steady-state, viscous incompressible flow is presented. Finite element equations are derived from a set of coupled nonlinear Navier-Stokes equations that consists of the conservation of mass and momentums. The convection terms in momentum equations are treated by streamline upwinding method to avoid the oscillation in the solution. The method has been developed for triangular element that employs equal-order interpolation functions for both the velocities and pressure. A segregated solution algorithm is also incorporated to compute the velocities and pressure separately. The method is combined with an adaptive meshing technique to further increase the solution accuracy, and at the same time, to minimize the computational time and computer memory requirement. The finite element formulation and the computer program have been verified by several examples that have known solutions prior to applying to solve more complex flow problems.

View full abstract

A supersonic blunt body problem was solved by the Euler equation, by using spreadsheets. Spreadsheets have an iteration function, and a finite difference Euler equation can be solved by flux vector splitting and the explicit time marching procedure. The cells in a spreadsheet are used as grids in computational fluid dynamics (CFD). We created computational space on spreadsheets based on standard CFD procedures. We first solved a blunt body problem, i. e., the flow around a sphere, with a prescribed approximate hyperbolic shock. This was an engineering approach and the result was satisfactory and comparable to that of the experiment. We used a coarse grid system and the shock fitting technique was subsequently employed by using the time-dependent cells. The asymptotic shock waveform was almost identical to the real one. The physical quantities of the flow were compared to the experimental results.

View full abstract

The molecular dynamics method for the interaction of gas molecules with a solid wall together with the Monte-Carlo method for the motion of gas molecules are applied to analyze the behavior of a slightly rarefied gas between two walls. The walls consist of platinum molecules, and the gas is taken to be xenon or argon. The Couette flow and the thermal problem for which two walls have different temperatures are considered. The slip and jump coefficients, tangential momentum and energy accommodation coefficients are obtained at the wall whose temperature is 300K. The tangential momentum accommodation coefficient of argon is as low as 0.19. The distribution functions of the reflected molecules are also obtained, and it is found that the Maxwell-type boundary condition describes well the distribution function of the reflected molecules.

View full abstract

The vortex structure in a separated flow region changes greatly owing to the introduced fluctuation in a mean flow. This study investigated experimentally the vortex structure in a separated flow. The periodic external fluctuation of velocity was introduced into the mean-flow using the disturbance generator set at the upstream of the separated region. In the experiment, two kinds of frequencies of the external fluctuation were chosen; the frequency measured in the separated shear layer and the frequency of vortex shedding without the external fluctuation. It was found that the separated region is reduced most effectively when the fluctuation frequency is multiples of the frequency of the vortex shedding of the separated region. The flow on the flat plate was found to become extremely periodic when the effective frequency fluctuation is introduced. This is due to the generation of the regular vortex structure on the flat plate.

View full abstract

Throat-mixing systems are proposed and assessed for the supersonic flow chemical oxygen-iodine laser (S-COIL). Three-dimensional, numerical simulation solving the governing equations of compressible gas flows together with chemical kinetics has been made for investigating the characteristics of the mixing condition and chemical reactions. The compressible Navier-Stokes equations and a chemical kinetic model encompassing 21 chemical reactions and 10 chemical species are solved by means of a full-implicit finite difference method. Two types of nozzles, a blade- and cylinder-type nozzles are adopted. The results show satisfactorily high values of the small signal gain coefficient G. The proposed throat-mixing system shows higher efficiency than the parallel mixing system. The blade-type nozzle is found to give 44% higher peak value of G than that of the parallel mixing system. It is also noted that the cylinder-type nozzle has superior ability than the parallel mixing system, in spite of its exceedingly simple structure.

View full abstract

Flow instability in a vaned diffuser with an even number of blades was examined experimentally and analytically. In the experiments, an alternate blade stall, an asymmetric stall, and two types of rotating stalls (backward/forward rotating stall) were observed depending on the impeller/diffuser clearance. For narrow clearance with strong impeller/diffuser interaction, the alternate blade stall and backward rotating stall mainly occurred. With increasing the clearance, the forward rotating stall also occurred, and the onset of rotating stall shifted toward the higher flow rate corresponding to the pressure performance in the vaned diffuser. Simple 2D stability analysis showed that the impeller/diffuser clearance affects the speed and direction of the stall propagation, and the slope of the diffuser pressure performance vs. flow rate curve affects fundamentally the onset of the flow instability within the diffuser.

View full abstract

In general, an unstructured grid technique has a good flexibility for a complex geometry, and already its usefulness has been demonstrated for full aircraft computations. In the case of three-dimensional, high Reynolds number viscous flow calculations, however, the very fine and stretched grids are required to resolve accurately thin boundary layers developed along the body surface, and so we often have some difficulties in high quality grid generation. A hybrid unstructured grid technique is incorporated into CASPER, a CFD-based design system developed at TRDI-JDA. In this paper, transonic and high Reynolds number flows around an ONERA Model M5 configuration are computed using the CASPER. To validate the present code, with respect to pressure distributions, longitudinal forces and moment coefficients, and transition lines, the present computed results are quantitatively compared with the other computed results and wind-tunnel testing data. Furthermore, some approaches are discussed for reliable drag prediction using CFD method.

View full abstract

Experimental studies are performed in order to clarify the effects of abrasive particle on the structure of the abrasive water suspension jet. Observations of jets are conducted at the injection pressure of 12MPa for seven types of abrasives: two types of steel bead, two types of alumina, two types of glass bead, and one type of plastic shot. The experiments show that the jet structure is greatly affected by the particle type and the concentration of the abrasive. By using smaller abrasive particles, the jet can be made more compact, whereas larger abrasive particles tend to promote jet breakup. These tendencies are noticeable in the case of high-density abrasive.

View full abstract

Motion and deformation of the wings of a moth Mythimna Separata (Walker) tethered to a steel beam were measured by using an optical method that uses fringe pattern projection. Simultaneously, the vertical force due to aerodynamic force generated by the wings was estimated by measuring the bending deformation of the beam during flapping motion of the moth. The force was estimated by subtracting the vertical forces due to inertial and centrifugal forces acting on the wings, which were estimated from the measured flapping motion, from the measured vertical force. Both the measured motion and deformation of the wing and the estimated vertical aerodynamic force generated by the wings suggest that the feathering angles are mainly affected by the inertial moment generated by the flapping motion and that the wing camber is mainly affected by the aerodynamic force generated by the wing.

View full abstract

Wake structures of a flat plate with finite thickness are studied using the rake of X-wires techniques. Parallel and oblique shedding modes in the wake flow are investigated. These modes depend on the flat plate end conditions. In the present paper, we demonstrate that manipulation of the end conditions yields vortex dislocations that appear at the same location. By manipulating the end conditions, vortex chain structures branching inside the dislocations are analyzed by means of pseudo-flow visualization of the X-wire rake data. The appearance of branching vortex chains is due to misfit between two oblique vortex shedding cells. Power spectra of velocity fluctuations show growth of the low-frequency mode due to subtraction of the two fundamental shedding modes of different frequency across the span. The small difference between the two fundamental shedding modes explains the large intermittent velocity irregularities in the large-scale structure downstream that occur in natural transition.

View full abstract

In this paper, we describe the fluid force acting on a two-dimensional oscillating circular cylinder in a lock-in region. The experiment was carried out in an N. P. L blow-down wind tunnel with a working section of 500mm×500mm×2000mm and Reynolds number of 1.9×10⁴. The cylinder was then forced to oscillate sinusoidally in the lift direction. The power spectrum of the fluctuating velocity in the wake behind the circular cylinder was measured to show the lock-in region in the present experiment. The time-mean pressure distribution and fluctuating pressure distribution on the circular cylinder were measured for the displacement in the oscillation. Consequently, it was found that the mean drag and fluctuating lift increase and become maximum in the lock-in region, while the base pressure at the rear surface (θ=180°) of the cylinder decreases and attains a minimum.

View full abstract

A simple physical model for small hermetic reciprocating, rotary and scroll compressors has been developed based on thermodynamic principles and large data sets from the compressor calorimeter and in situ tests. Pressure losses along the refrigerant path are neglected and a compression process is assumed isentropic. A mass flow rate model reflects clearance volumetric efficiency and simulates suction gas heating using an effectiveness method. Compressor work is calculated using the compressor efficiency represented by only two empirical parameters. A linear relationship between the discharge and shell temperatures is extracted from the large data sets and applied to the model for calculating the discharge temperature. Another experimental observation indicates that the specific volumes at the suction and discharge ports of the cylinder have linear relationships with the specific volumes at the compressor suction and discharge. Those relationships can be used for the compressor model and the possibility is reported. The models developed using physical principles and experimental observations can predict the mass flow rate and power consumption within ±3.0% accuracy.

View full abstract

A computational technique for the two-dimensional non-equilibrium homogeneously condensing flows through steam turbine cascades is presented. The fundamental equations of compressible wet steam flows are descretized based on the TVD scheme and following thermodynamic assumptions. 1) The gas phase is an ideal gas. 2) A liquid phase consists of droplets whose radii are in the order of 10^-6m or less, thereby, the wet steam is a homogeneous fluid. 3) A gas-liquid phases are changed by the homogeneous nucleation and growth of the existing droplets, which are described by the classical nucleation theory. In this theory, a driving force of a phase change is a degree of supercooling. The described assumptions are valid for the flow through the low-pressure stage of the steam turbines. Flow through the moving blades for a low-pressure steam turbine is calculated. The calculated pressure distributions on the blade surface agree well with experimental data. The influences of both the degree of inlet supercooling and the ratio of the inlet total pressure to outlet static pressure on the blade performance, mass flow, flow angle, energy loss are studied.

View full abstract

An ignition and combustion model has been developed to predict the heat release rate in direct injection compression ignition engines employing very early injection timing. The model describes the chemical reactions, including low-temperature oxidation. The KIVA II computer code was modified with the present ignition and combustion model. The numerical results indicate that the model developed in this work reproduces major features of two-stage autoignition, as well as experimentally observed trends in NO_x and unburned fuel emissions. The computational results show that fuel injection timing significantly influences NO_x emissions. Results also indicate that fuel droplets that enter the squish region possibly become unburned fuel emissions. Some graphical results demonstrate the relationships among the in-cylinder fuel spray distributions, fuel-air equivalence ratio, temperature, and mass fractions of NO and unburned fuel.

View full abstract

In order to improve steam turbine internal efficiency, a new highly loaded rotor blade for the high and intermediate pressure (HIP) stages, has been developed. The blade configuration is expressed by a 5th-order Bezier curve and an interactive technique is adopted in combination with a 2-D turbulent blade-to-blade flow analysis. The inverse method is adopted to modify the blade shape by optimizing the blade aerodynamic loading distribution near the root section which should decrease not only profile loss, but also endwall loss. Three-dimensional stage flow analysis and air turbine tests are carried out to verify the stage performance of the new blade. It is found that the new blade has improved blade performance near the root and the midspan; in particular, higher blade performance near the midspan contributes to an improvement in stage efficiency by about 0.3% at the design point. Consequently, it is verified that the new rotor blade is effective to improve steam turbine internal efficiency and to lower manufacturing cost by reducing the blade number by 15%.

View full abstract

A basic characteristics of compressed natural gas direct-injection (CNG DI) combustion was studied by using a rapid compression machine. Results show that comparing with homogeneous mixture, CNG DI has short combustion duration, high pressure rise due to combustion, and high rate of heat release, which are considered to come from the charge stratification and the gas flow generated by the fuel injection. CNG DI can realize extremely lean combustion which reaches 0.03 equivalence ratio, φ. Combustion duration, maximum pressure rise due to combustion and combustion efficiency are found to be insensitive to the injection modes. Unburned methane showed almost the same level as that of homogeneous mixture combustion. CO increased steeply with the increase in φ when φ was greater than 0.8 due to the excessive stratification, and NO_x peak value shifted to the region of lower φ. Combustion inefficiency maintains less than 0.08 in the range of φ from 0.1 to 0.9 and increases at very low φ due to bulk quenching and at higher φ due to excessive stratification. The combustion efficiency estimated from combustion products shows good agreement with that of heat release analysis.

View full abstract

This study deals with the exchange flow of two different gases(He-air, Ar-air, and SF₆-air)through a rectangular channel that has a 5-mm width, a 50-mm height, and a 200-mm length. The net exchange flow rate is measured by an electronic mass balance, and velocity distribution is measured by a laser-Doppler anemometer. Flow patterns of the exchange flow are made visual with a tracer method using smoke. The effects of gas densities, molecular diffusion coefficients, inclination angles (θ) of the channel, on the flow patterns and the net exchange flow rates, are discussed. Based on the experimental results, the net exchange flow data are correlated by the equation derived by dimensional analysis: Q^*=2.87×10^-7Gr^2.17Sc^0.41θ^-1.09.

View full abstract

This paper proposes a novel technique that can augment heat transfer but nevertheless can reduce pressure-loss in a fin-tube heat exchanger with circular tubes in a relatively low Reynolds number flow, by deploying a single row of the delta winglet type of vortex generators beside the front row of the three-row tube bundles in an in-line arrangement. The winglets are placed with a heretofore-unused orientation for the purpose of augmentation of heat transfer. This orientation is called as “common flow up” configuration. The proposed device causes a significant separation delay, reduces form drag, and removes the zone of poor heat transfer from the near-wake of the tube. This enhancement strategy has been successfully verified by experiments in the proposed configuration. In the in-line tube bundle with three rows, the heat transfer was augmented by 10% to 20%, and yet the pressure loss was reduced by 8% to 15% for the Reynolds number (based on two times channel height) ranging from 350 to 2100, when the present winglets were built in a single front row.

View full abstract