In this paper, the current status of the hypersonic research is reviewed, and the areas needing further work are proposed

A homogeneous flow field including more than 2000 spherical particles was directly simulated. Particles are settling by gravity with the Reynolds number ranging from 50 to 300, based on diameter and slip velocity. Particular attention was focused on the distribution of particles. The Reynolds-number dependence, influences of particle rotation and loading ratio, and the dynamics of particle clusters are discussed. In the higher Reynolds number case, the wake attraction causes particle clusters and the average drag coefficient decreases significantly. Non-rotating particles maintain cluster structure and rotating ones moves randomly in the horizontal direction. It is because of the difference in the direction of the lift force.

Two different ${\kappa}-{\varepsilon}-\overline{{\nu}{\nu}}-f$ turbulence models together with the two-layer model are evaluated for natural convection in a rectangular cavity. The numerical problem and accuracy of the turbulence models are discussed. The original $\overline{{\nu}{\nu}}-f$ model suffers from the numerical stiffness problem when used with the segregate solution procedure like the SIMPLE algorithm, and a remedy for this problem is proposed. It is shown that original $\overline{{\nu}{\nu}}-f$ model best predicts the mean velocity, Reynolds stresses and the turbulent heat flux while the modified $\overline{{\nu}{\nu}}-f$ model (N=6) overpredicts the turbulent quantities.

An implicit pressure correction method on unstructured Cartesian grid is developed for the incompressible Navier-Stokes equations. An immersed boundary method is also incorporated to treat the body geometry. Tests show that with an appropriate amount of dissipation, the method is second order accurate both in time and space. The driven cavity flows with and without immersed bodies are computed to demonstrate the capability of the present scheme.

Drop deformation in a cylindrical contraction geometry, characterized here by the formation of a thread of drop fluid in the contraction, is predicted using a Volume-of-Fluid numerical technique. The predicted drop shape is found to closely follow the observed deformation. A capillary instability in the developing drop thread in the contraction was predicted, in agreement with experimental observation.

Brownian-dynamics simulation on highly charged colloidal suspensions is performed by employing Tokuyama effective force recently proposed. The radial distribution function suggests that there exist three novel phases, a gas phase, a liquid droplet phase, and a face-centered cubic (FCC) crystal droplet phase, depending on the minimum values of that potential. The dynamics of droplet growth is also investigated both in liquid droplet phase and in crystal droplet phase. Thus, different types of characteristic growth stages are found.

We perform two kinds of computer simulations on polydisperse hard-sphere systems; a molecular-dynamics simulation on atomic systems and a Brownian-dynamics simulation on colloidal suspensions. Analyses of the mean square displacement, the radial distribution function, and the pressure suggest that there exist three phase regions, a liquid phase region, a metastable phase region, and a crystal phase region, where the freezing and melting points are shifted to the values higher than in monodisperse case. It is also shown that the long-time behavior of colloidal suspensions is exactly the same as that of atomic systems.

It is shown that using the unified coordiantes of Hui et al.[1 - 4], one can now compute fluid flow without prior grid generation. This represents a great saving of computing time.

Three-dimensional numerical analysis of the flow around rectangular cylinders with various side ratios, D/H, from 0.2 to 2.0 is carried out for Reynolds number of 1000 by using multi-directional finite difference method in multi-grid. The predicted results are well compared with the experimental data. It is found that fluid dynamics characteristics alternate between high pressure mode. and low pressure mode of the base pressure for rectangular cylinder of D/H=0.2-0.6.

n this paper, direct numerical simulation of decaying compressible turbulence with passive scalar is performed by using 7th order upwind difference scheme or 8th order group velocity control scheme. The start Reynolds number (defined by Taylor scale) is 72 and turbulent Mach numbers are 0.2-0.9. The Schmidt numbers of passive scalar are 2-10. The Batchelor k-1 range are found in scalar spectra, and the high wavenumber spectra decays faster with increasing turbulent Mach number. The extend self-similarity (ESS) is found in the passive scalar in compressible turbulence.

Nonlinear relationship between Reynolds stresses and the rate of strain for nonlinear${\kappa}-{\varepsilon}$ turbulence models is validated theoretically by using the boundary layer assumptions against the turbulence­driven secondary flows in noncircular ducts and then the prediction performance for several nonlinear models is evaluated numerically through the application to the turbulent flow in a square duct.

Direct numerical simulations are performed for a turbulent flow subjected to a sudden change in pressure gradient. The calculations are started from a fully-developed turbulent channel flow at $Re_{\tau}=180$. The pressure gradient of the channel flow is then changed abruptly. The responses of the turbulence quantities (e.g., turbulence intensities, Reynolds shear stress, and vorticity fluctuations) and the near-wall turbulence structure to the pressure gradient change are investigated. It is found that there are two different relaxations: a fast relaxation at the early stage and a slow one at the later stage. The early response of the velocity fluctuations shows an anisotropic response of the near-wall turbulence.

In the present research a high order Gudonov-type method has been used for the simulation of very high pressure flow fields, as well as the capturing of strong shocks, which usually occur in explosion of high explosives. The treatment strong shocks and the flow field behind the shocks needs a very high resolution scheme. To resolve accurately the shock and the release waves behind the shock the piece­wise parabolic method (PPM) of Colella [1] was utilized in this research. A major problem which encountered in very high pressure problems is the equation of state which differs completely form the ideal-gas equation of state (EOS). Here, the original PPM is extended for real gas effect consideration.

A numerical solution is presented for the natural convection heat transfer from two vertical fins using a spectral finite difference method. Virtual distant boundary conditions for two bodies that are compatible with plume behavior and with an overall continuity condition are introduced. A boundary-fitted coordinate system is formed. Streamlines, isotherms, mean Nusselt numbers and drag & lift coefficients are presented for a variety of dimensionless parameters such as a Grashof number and a Prandtl number at a steady-state. Extensive effectiveness of a spectral finite difference method was established.

In this paper, the new higher order wall boundary conditions are proposed for solving the incompressible flows. The square driven cavity flows are simulated by using the variable order method of lines with the present wall boundary conditions. The variable order method of lines is constructed by the spatial discretization, i.e., the variable order proper convective scheme for convective terms and the modified differential quadrature method for diffusive terms, and time integration. The 2nd, 4th, and 6th order solutions are presented and these results show this higher order boundary conditions are very promising for the incompressible flow simulations.

General classes of boundary-pressure-driven flows of incompressible Newtonian fluids in three­dimensional (3D) channels and in 3D pipes with known steady laminar realizations are investigated respectively. The characteristic physical and geometrical quantities of the flows are subsumed in the kinetic Reynolds number Re and a parameter $\psi$, which involves the energetic ratio and the directions of the boundary-driven part and the pressure-driven part of the laminar flow. The solution of non-stationary dimension-free Navier-Stokes equations is sought in the form $\underline{u}=u_{L}+U,\;where\;u_{L}$ is the scaled laminar velocity and periodical conditions are prescribed for U in the unbounded directions. The objects of our numerical investigations are autonomous systems (S) of ordinary differential equations for the time-dependent coefficients of the spatial Stokes eigenfunction, where these systems (S) were received by application of the Galerkin-method to the dimension-free Navier-Stokes equations for u.

A CFD analysis has been made for fully developed turbulent flows in a triangular bare rod bundle with pitch to diameter ratio (P/D) of 1.123. The nonlinear turbulence models predicted the turbulence­driven secondary flow in the triangular subchannel. The nonlinear quadratic $\kappa-\omega$ models by Speziale and Myong-Kasagi predicted turbulence structure in the rod bundle fairly well. The nonlinear quadratic and cubic $\kappa-\omega$ models by Shih et al. and Craft et al. showed somewhat weaker anisotropic turbulence. The differential Reynolds stress model appeared to overpredict the turbulence anisotropy in the rod bundle.

Direct numerical simulation results of aeolian tones generated by a two-dimensional obstacle (circular cylinder, square cylinder, NACA0012 airfoil) in a uniform flow are presented and the generation and propagation mechanisms of the sound are discussed. The unsteady compressible Navier-Stokes equations are solved by a highly-accurate finite difference scheme over the entire region from near to far fields. The direct numerical simulation results are also compared with the results obtained by Curle's acoustic analogy.

The stability of flows induced by a surface acoustic wave (SAW) propagating along the deformable walls in a confined parallel-plane microchannel or slab in the laminar flow regime is investigated. The governing equation which was derived by considering the nonlinear coupling between the deformable or waving interface and viscous fluids is linearized and then the problem is solved by a verified code based on the spectral method together with the associated interface and boundary conditions. The value of the critical Reynolds number was found to be near 1439 which is much smaller than the rigid-wall case: 5772 for conventional pressure-driven flows.

A parallelized FEM code based on domain decomposition method has been recently developed for a large scale computational fluid dynamics. A 4-step splitting finite element algorithm is adopted for unsteady computation of the incompressible Navier-Stokes equation, and Smagorinsky LES(Large Eddy Simulation) model is chosen for turbulent flow computation. Both METIS and MPI library are used for domain partitioning and data communication between processors respectively. Tiburon of Hyundai-motor is chosen as the computational model at $Re=7.5{\times}10^{5}$, which is based on the car height. It is confirmed that the drag under road condition is smaller than that of wind tunnel condition.

Unstructured grid system is suitable for flows of complex geometries. For problems with moving boundary walls, the grid system must be changed and deformed with time if we use a body fitted grid system. In this paper, a new moving-grid finite-volume method on unstructured grid system is proposed and developed for unsteady compressible flows with shock waves. To assure geometric conservation laws on moving grid system, a control volume on the space-time unified domain is adopted for estimating numerical flux. The method is described and applied for two-dimensional flows.

We implement Realtime Air Diffusion Prediction System which is designed for air diffusion simulations with four-dimensional data assimilation. For realtime running, we parallelize the system using MPI (Message Passing Interface) on distributed-memory parallel computers and build a cluster computer which links high-performance PCs with high-speed interconnection networks. We use 162­CPU nodes and a Myrinet network for the cluster

The spectral eddy viscosity model is investigated through the large eddy simulation of the decaying and forced isotropic turbulence. It is shown that the widely accepted 'plateau and cusp' model overpredicts resolved kinetic energy due to the amplification of energy at intermediate wavenumbers. Whereas, the simple plateau model reproduces a correct energy spectrum. This result overshadows a priori tests based on the filtered DNS or experimental data. An alternative method for the validation of subgrid-scale model is discussed.

We propose a wall distance free one-equation turbulence model. The model is organized in an extremely simple form. Only a few model constants were introduced into the model. The model is numerically tough and easy-of-use. The model also demonstrated the ability to simulate the laminar to turbulent flow transition. The model has been applied to the channel flow, the plane jet, the backward facing step flow, the flat plate boundary layer, as well as the flow around the 2D airfoil at large angles of attack, which obtained satisfactory results.

A numerical simulation of an incompressible cavity flow is conducted using turbulence models. Cavity geometry and flow conditions are based on Cattafesta's experiment. Baldwin-Lomax model and ${\kappa}-{\varpi}$ model are employed. While simulation with Baldwin-Lomax model predicts the oscillatory features of the flow, the use of ${\kappa}-{\varpi}$ model in its original form makes the simulation converge to steady flow. To acquire oscillatory flow solution, Kato-Launder form and Time scale bound are adopted in production term of ${\kappa}-{\varpi}$ model. The strouhal number of the flow oscillations from the simulation results corresponds to 1 st mode in simulation but 2 nd mode in experiments. However mean velocity profile is in good agreement with the experimental data and the fluctuation profile follows the tendency of Cattafesta's results.

Droplet impinging into a cavity at micro-scale is one of important fluidic issues for microfabrications, e.g. bio-chip applications and inkjet deposition processes in the PLED panel manufacturing. The droplets generally dispensing from an inkjet head, which contains an array of nozzles, have a volume in several picoliters, while each nozzle jets the droplets into cavities with micron-meter size located on substrates. Due to measurement difficulties at micro-scale, the numerical simulation could serve as an efficient and preliminary way to evaluate the micro-sized droplet impinging behavior into a cavity. The micro-fluidic flow is computed by solving the three-dimensional Navier-Stokes equations through a finite volume discretization. The droplet front is predicted by a volume-of-fluid approach, in which the surface tension is modeled as a function of the fluid concentration. This paper discusses the influence of fluid properties, such as surface tension and fluid viscosity, on micro-fluidic characteristics at different jetting speeds in the deposition process via the proposed numerical approach.

The behavior of viscous fingerings caused by an external force is investigated by using a two­phase lattice Boltzmann method. The effects of the modified capillary number, the viscosity contrast, and the modified Darcy-Rayleigh number on the instability of interfaces are found. The calculated wave numbers are in good agreement with the theoretical ones in the range of wave numbers smaller than 10, but the calculated ones tend to become smaller than the theoretical ones in higher wave numbers.

In this study a microchannel with various arrangement of blocks is newly proposed. This design comprises periodically arranged simple blocks. In this configuration, the stirring is greatly enhanced at a certain geometric parameter set. To characterize the flow field and the stirring effect both the numerical and experimental methods were employed. To obtain the velocity field, three-dimensional numerical computation to the Navier Stokes equations are performed by using a commercial code, FLUENT 6.0. The fluid-flow solutions are then cast into studying the characteristics of stirring with the aid of Lyapunov exponent. The numerical results show that the particles' trajectories in the microchannel heavily depend on the block arrangement. It was shown that the stirring is significantly enhanced at larger block-height and it reaches maximum when the height is 0.8 times the channel width. We also studied the effect of the block stagger angle, and it turns out that the stirring performance is the best at the block angel ${45^\circ}$.

Understanding the physics of electroosmotic flow is important to the optimum design and operation of microfluidic systems. Recent development on the topic in our group is summarized in this presentation. Applications to the variable-volume-Injection, micro flow switch, electrokinetic focusing, and flow cytometer will be presented with our novel design and control.

In the generalized characteristic coordinate system (GCCS) proposed by Wu and Shi [1], the frame moves at a speed which is a linear combination of the convective speed and the sound speed, thus unifying the classical Eulerian approach, Lagrangian approach, and the unified coordinate system (UCS) of Hui and his co-workers [2]. Here some properties of Euler equations in the GCCS are studied and the advantages of GCCS in capturing expansion fans and shock waves are demonstrated by the results of numerical tests.

The streamfunction-vorticity equations for two-dimentional cavity flow are solved by a new finite element method which uses finite spectral basis functions as interpolation functions for rectangular elements. Results for several cases with different Renold's number are compared with benchmark solutions and found to be in well agreement.

As a part of numerical and experimental research works for the prediction and improvement of ship's maneuvering performance, a study on the performance analysis of two different rudders has been carried out. While the planform shape and the aspect ratio of the rudders have been fixed, section shape has been changed. Conventional type of HMRI NP section and special type of dolphin-tail section have been employed. Performances of the rudders have been investigated by using CFD and compared with experimental data obtained in a wind tunnel. A commercial CFD program has been used to solve the RANS equations. Two-equation k-ro model has been applied to close the governing equations. Block-structured grids are used in the numerical calculation. Based upon the calculation results, the rudder with dolphin-tail section has shown a possibility of significantly improving rudder performance if utilized as the section of ship rudders.

When the under-expanded supersonic jet impinges on the perpendicular plate, it is well known that the self-induced flow oscillation occurs at the specific conditions. This phenomenon is related with the noise problems of aeronautical and other industrial engineering. But, the very complicated flow field is formed and it is difficult to clear the flow structure and the mechanism of oscillation. This paper aims to clear the characteristics of flow field and the wave pattern during the under-expanded supersonic jet impinges on the plate. The numerical calculation was carried out using the TVD numerical method. In this paper, the flow visualization, the pressure fluctuation on the surface of plate and the mechanism of oscillation are discussed.

A feasibility study on the alternating jet flow under the static electromagnetic field was carried out. When a fluid with electrical conductivity lies in the static electromagnetic field and moves electric current occurs in the fluid. Due to the electromagnetic field and the electric current, lorentz force generates in the fluid, which undergo the 'breaking' effect to the fluid. In order to simulate the complex fluid flow in the magnetic field, electromagnetic and fluid flow analysis need to be solved simultaneously. In the present study, a SOLA (SOLution Algorithm) scheme was used in order to calculate electromagnetic and fluid flow field. Jet flow without an electromagnetic field was compared with analytical solution in order to validate the flow analysis scheme. Effect of jet velocity on the flow pattern down the jet was investigated.

Using computational fluid dynamics with the fluid-structure interaction, structural effects of intra-luminal thrombus were determined in thrombosed axisymmetric abdominal aorta aneurysm (AAA) models under pulsatile flow. Four different models, varying dilatations of the aneurysm and Young's moduli of intra-luminal thrombus, were defmed. Compared with unthrombosed AAA models, both von Mises stress and radial displacements in the aneurysm wall significantly decreased. Stiffer intra-luminal thrombus reduced von Mises stress in the aneUtysm wall. The present study supported that intra-luminal thrombus might reduce wall stress in the aneurysm.

A simple method is proposed to split the flux vector of the Euler equations by introducing two artificial wave speeds. The direction of wave propagation can be adjusted by these two wave speeds. This idea greatly simplifies the upwinding, and leads to a new family of upwind schemes. Numerical flux function for multi-dimensional Euler equations is formulated for any grid system, structured or unstructured. A remarkable simplicity of the scheme is that it successfully achieves one-sided approximation for all waves without recourse to any matrix operation. Moreover, its accuracy is comparable with the exact Riemann solver. For 1-D Euler equations, the scheme actually surpasses the exact solver in avoiding expansion shocks without any additional entropy fix. The scheme can exactly resolve stationary contact discontinuities, and it is also freed of the carbuncle problem in multi­dimensional computations.

Thermocapillary convection is a surface tension driven flow due to a temperature gradient along an interface. It occurs during a crystal-growth process and therefore understanding the convection is important to material processing in microgravity. Although modelling of the float-zone crystal-growth process has been of interest for a few decades, most studies of liquid bridges assumed non-deformable flat surfaces. In reality, the surface profile, g(t,z), is unknown and should be obtained as a solution to the coupled transport equations along with the surface force balance. Here we report on a numerical study of axisymmetric thermocapillary convection in liquid bridges with deformable surfaces. The interface is determined as part of the complete solution. The influence of the capillary number (Ca), Reynolds number (Re), Prandtl number (Pr) and aspect ratio(Ar) on the dynamics is explored.

The problem of a bluff body oscillating in a fluid flow has been receiving a great deal of attention. When a bluff body is placed in a flow, it experiences fluctuating hydraulic forces in both transverse and stream-wise directions. It is caused by the formation of vortices behind the body, which could cause large damages of structures. It is called the flow-induced vibrations. In this article, it is investigated the effects of that side-jets and rear-jet, which is applied to control the vortex shedding. The rear-jet is available to control the flow-induced vibrations according as the body shapes and the velocity of fluid flow in which the galloping phenomena is not appeared.

This paper presents a theoretical study concerning multibubble dynamics in a sound field and the numerical validation for it by employing our new CFD code MTS-DiCUP. In recent papers, the author has shown theoretically that an unknown characteristic frequency, named 'transition frequency,' exists in a multibubble system. For a N -bubble case, up to 2N -1 transition frequencies per bubble have been predicted, only N ones of them correspond to the natural frequencies of the system. The transition frequencies that do not correspond to the natural frequencies give rise to the phase reversal of bubbles' pulsation without resonant response. In this paper, it has been suggested theoretically that those transition frequencies may cause the sign reversal of the secondary Bjerknes force, which is an interaction force acting between acoustically coupled gas bubbles. This theoretical result has been validated by the direct numerical simulation, at least in a qualitative sense.

Nnumerical investigations of the tip vortical characteristics were conducted with lateral tip blowing to reduce Blade-Vortex Interaction (BVI) noise. The predictions of BVI noise were performed using a combined method of an unsteady Euler code with an aeroacoustic code based on Ffowcs- Williams and Hawkings formulation. A moving overlapped grid system with three types of grids (blade grid, inner and outer background grid) was used to simulate BVI of helicopter with two OLS-airfoil blades in forward/ descending flight condition. The calculated waveform of BVI noise, which is characterized by the distinct peaks caused during blade vortex interaction, clearly shows the effect of lateral blowing at tip to reduce BVI noise

The convergence characteristics of the LV scheme for the Euler equations have been investigated by using the Von Neumann stability analysis. The results indicated that the convergence rate is governed by a specific combination of CFD parameters. Based on this insight, it is shown that the convergence characteristics of the LV scheme is not deteriorated at any grid aspect-ratio as long as the local time step is defined based on the parameter combination. The numerical results demonstrated that this time step definition provide a uniform convergence for grid aspect-ratios between one to$1{\times}10^{4}$.

A preconditioned numerical method for gas-liquid to-phase flow is applied to solve cavitating flow. The present method employs a density based finite-difference method of dual time-stepping integration procedure and Roe's flux difference splitting approximation with MUSCL-TVD scheme. A homogeneous equilibrium cavitation model is used. The method permits simple treatment of the whole gas-liquid two-phase flow field including wave propagation, large density changes and incompressible flow characteristics at low Mach number. By this method, two-dimensional internal flows through a venturi tuve and decelerating cascades are computed and discussed.

In order to reduce numerical dissipation in a multi-dimensional flow, a new variable interpolation method for Roe's FDS is proposed. By introducing the Mach number weighting function w, the properties at the cell-interface are interpolated and in a non flow-aligned grid system, it can give more accurate solution because of less numerical dissipation. Various test cases including oblique contact discontinuity are simulated and compared with the results of original Roe's FDS.

In order to control the transonic flow field with shock wave, a condensing flow was produced by an expansion of moist air on a circular bump model and shock waves were occurred in the supersonic parts of the fields. Furthermore, the additional passive technique of shock - boundary layer interaction using the porous wall with a cavity underneath was adopted in this flow field. The effects of these methods on the shock wave characteristics were investigated numerically. The result showed that the flow fields might be effectively controlled by the suitable combination between non-equilibrium condensation and the position of porous wall.

A New kinetic scheme for the compressible Navier-Stokes equations is developed. While the conventional approach, such as KFVS scheme, employs the splitting algorithm and computes the numerical flux on the basis of the collisionless equation, the present approach employs the splitting algorithm in the evaluation of the numerical flux, where the collision effect is explicitly taken into account. However, the initial condition employed in the computation is slightly different from the conventional Chapman-Enskog NS distribution function. The present study also reveals the background of the existing kinetic schemes. such as the KFVS scheme and Gas-Kinetic BGK scheme.

A numerical method for a wave diffraction problem in three-dimensional channels is developed. The physical models are various shapes of channel connected to the open sea. When a ship or an offshore structure is moored in various configurations of channel connected to an open sea, the prediction of the hydrodynamic force exerting on the moored ship could be important for the prediction of its motion. It is assumed that the fluid is inviscid and incompressible and its motion is irrotational. From the continuity equation, the Laplace equation can be obtained as the governing equation. The surface tension at free surface is neglected, and wave amplitude is assumed to be small compared to the wave length. Then the free surface condition can be linearized. The numerical method used here is the localized finite element method based on a variational formulation

In this work, the Greitzer's B-parameter model is applied for analyzing the stall and surge characteristics. The four parameters in the model are highlighted in order to establish the influence of each parameter on the system. First of all, the governing equations of stall and surge behavior are solved numerically using fourth-order Runge-Kutta method. The Taguchi method is then used to analyze the results generated to obtain the extent of effects of the parameters on the system by varying the parameters in a series of combinations. Finally, a thorough analysis is carried out on the results generated from the Taguchi method and the graphs.

In this paper, Gauss filter function is used to filter the N-S equation and the subgrid-scale Reynold stresses model is introduced to deduce the practical form of LES equation for 2-D case for flow calculation of hydraulic machine. Then the LES equation and its discrete form in computational field are obtained in the body-fitted coordinate system and the numerical calculation program is built. The incompressible turbulent flow in double-flow-conduits-sewage pump impeller is computed by using the abovementioned program, and then the distribution rules of velocity and pressure in flow field are obtained. Based on this, the designs of double-flow-conduits-sewage pump impeller are optimized.

Fly ash enters axial compressor when a turbomachinery is operated in an adverse environment. We have numerically investigated erosion of the blade and shroud in the turbulent compressor passage flow under the influence of gas-particle two-phase interaction. There have appeared quasi-three dimensional calculations on this subject but not the complete three-dimensional gas-particle interaction as done in the present work. Lagrangian particle tracing technique is used on the base of parallel processing for efficient calculation. Accuracy of the present code is tested using the benchmark JPL nozzle. In the DFVLR compressor blades, we have shown that a large number of particles passing through the tip clearance make impact on the blade tip and on the shroud. Higher degree of erosion is resulted by the heavier particles due to the centrifugal force.

Increasing interest in indoor air quality (IAQ) control has been found because of its serious effect on human health. To evaluate IAQ, thermal comfort in terms of temperature and velocity distributions of indoor air has to be analyzed in detail. Choice of location for installation of air-conditioner in a building will affect the performance of cooling effect and thermal comfort on the occupants, which in turn will affect the indoor air quality (IAQ) of the building. In this paper, we present a discussion on the proper location of the air-conditioner in order to obtain good thermal comfort for occupant of a typical bedroom in Macao. A set of carefully designed numerical experiments is run with the Computational Fluid Dynamics (CFD) software FLOVENT 3.2 [1]. Reynolds averaged Navier-Stokes equations are solved with finite volume technique and turbulence effects upon the mean flow characteristics is modeled with the k - & model. Assumption of steady state environment is made and only convective and conductive heat transfer from the occupant and air-conditioner are being concerned.

Three dimensional flow above a sand dune has been studied numerically by using Large-Eddy Simulation (LES) method. The movement of the sand which is formed by converging wind directions has been investigated. The numerical method employed in this study can be divided into three parts: (i) calculation of the air flow above the sand dune using MAC method with a generalized coordinate system; (ii) estimation of the sand transfer caused by the flow through the friction; (iii) determination of the shape of the sand surface. Since the computational area has been changed due to step (iii), (i)­(iii) are repeated. The simulated dune, which has initially elliptic cross section, extending at the converging direction which is known as linear dunes.

This paper put forward the controlled spin intensity method for the tangentially-fired furnaces to solve the problems existed in the counter-tangential operation. The numerical simulation was used in this paper to discuss some basic principles.

The tidal bores of the Qiantang River on the East coast of China are simulated numerically based on the shallow water theory. The governing equations, which were traditionally formulated using water depth, are formulated in terms of water surface level, and the fractional-step method is applied in conjunction with a Godunov-type scheme. In addition, the source terms due to bottom gradient are discretized centrally to exactly balance the flux terms. Our numerical simulation produces tidal bores in excellent agreement with field measurements.

Complex flow field around the Darius turbine rotating stationally are simulated by solving the three dimensional incompressible Navier-Stokes equation numerically. The rotating coordinate system is employed so that the boundary conditions on the blades of the rotor become simple. In order to impose the boundary condition on the blades precisely, the boundary fitted coordinate system is employed. Fractional step method is used to solve the basic equations. The complex flow fields due to the three dimensionality of the geometry of the turbine and the rotation of the turbine are obtained and they are visualized effectively by using the technique of the computer graphics.

Two and three-dimensional flows around a cross-flow wind turbine are investigated by the numerical simulation. The turbine studied in this paper has cylindrical shape with many small blades along its periphery. Incompressible Navier-Stokes equation is used for this simulation. A rotating coordinate system, which rotates at the same speed of the turbine, is used in order to simplify the boundary conditions on the blades of the turbine. Additionally, a boundary fitted coordinate system is employed in order to express the shape of the blades precisely. A third order upwind scheme is chosen for the approximation of the non-linear terms. When the number of blades is about 10, the highest torque is obtained.

Fly ash enters axial compressor when a turbomachinery is operated in an adverse environment. We have numerically investigated erosion of the blade and shroud in the turbulent compressor passage flow under the influence of gas-particle two-phase interaction. There have appeared quasi-three dimensional calculations on this subject but not the complete three-dimensional gas-particle interaction as done in the present work. Lagrangian particle tracing technique is used on the base of parallel processing for efficient calculation. Accuracy of the present code is tested using the benchmark lPL nozzle. In the DFVLR compressor blades, we have shown that a large number of particles passing through the tip clearance make impact on the blade tip and on the shroud. Higher degree of erosion is resulted by the heavier particles due to the centrifugal force.

The present study investigates on design optimization of rib-roughened two-dimensional channel to enhance turbulent heat transfer. Response surface method with Reynolds-averaged Navier-Stokes analysis is used as an optimization technique. Standard $k-{\varepsilon}$model with wall functions is adopted as a turbulence closure. The objective function is defined as a linear combination of heat transfer and friction drag coefficients with weighting factor. Computational results for overall heat transfer rate show good agreements with experimental data. Four design variables are optimized for weighting factor of 0.02.

A method for performing two-dimensional lift-constraint drag minimization in inviscid compressible flows on unstructured meshes is developed. Sensitivities of objective function with respect to the design variables are efficiently obtained by using a continuous adjoint method. In addition, parallel algorithm is used in multi-point design optimization to enhance the computational efficiency. The characteristics of single-point and multi-point optimization are examined, and the comparison of these two method is presented.

Large Eddy Simulation(LES) of turbulent spray combustion flow was conducted. An experimental database for the laboratory spray combustor is chosen to validate the present numerical simulation. The governing equations for the gas phases are discretized in three-dimensional curvilinear boundary-fitted coordinate system, and the fuel droplet motion equations are described in Lagrangian representation. The numerical results are compared with the experiment for the gas-phase mean velocities and its fluctuation in cold flow condition. Three dimensional vortical structures are well visualized and droplet motion is well predicted.

Thermal analysis is performed to protect the propulsion system of low-earth-orbit earth observation satellite from unwanted thermal disaster like propellant freezing. To implement thermal design adequately, heater powers for the propulsion system estimated through the thermal analysis are decided. Based on those values anticipated herein, the average power for propulsion system becomes 22.02 watts when the only one redundant catalyst bed heater is turned on. When for the preparation of thruster firing, 25.93 watts of the average power is required. All heaters selected for propulsion components operate to prevent propellant freezing meeting the thermal requirements for the propulsion system with the worst-case average voltage, i.e. 25 volts.

Computations of the mean and turbulence flows over three-dimensional hill of conical shape have implemented. Beside the standard ${\kappa}-{\varepsilon}$ , two other modifications proposed by Detering & Etling and Duynkerke for atmospheric applications were also considered. These predictions were compared with the data of a wind tunnel experiment. From the comparison, it was concluded that all three models predict the mean flow velocities equally well while only the Duynkerke's model accurately predicts the turbulence data statistics. It also concluded that there are large discrepancies between model predictions and the measurements near the ground surface. The flow field, which was obtained by using the Duynkerke's modification, was used to simulate gas dispersion from an upwind source. The calculation results are verified based on the measurement data. Modifications of the turbulent Schmidt number were carried out in order to match the measured results. The code was used to investigate the influence of the recirculation zone behind a building of cubical shape on the transport and dispersion of pollutant. For a stack behind and near the obstacle, some conclusions about the effect of the stack height and stack location were derived.

Simulations of atmospheric diffusion process under stable and unstable conditions were carried out using both numerical and experimental methods. Results from the previous study show that numerical simulation using 3-dimensional incompressible Navier-Stokes equation and density deviation are in good agreement with typical plume pattern. In this study, we use experimental data of temperature and wind profile obtained from a thermally stratified wind tunnel as initial conditions for numerical simulation and compare the results.

Flow around Savonius rotor is studied by means of the numerical simulation. Three-dimensional incompressible Navier-Stokes equations are solved numerically. Overgrid system is employed in order to enable the flow calculation of complex geometry. The basic equations in each region are solved by using the standard MAC method. The physical quantities such as the velocity and the pressure among each region are transferred through the overlapping region which is common in each region. Some numerical results of static and rotating rotor will be presented.

Cross-flow fans are widely used in various applications, due to their large capacity of mass flow and size compactness. The flow field of the cross-flow fan is, however, complex and has many design parameters. Thus, the general design guide has not been sufficiently established yet and the design strategies of cross-flow fans have been based on experiments. In the present study, the performance and their two-dimensional flow characteristics are numerically analyzed by using the STAR-CD(commercial computational fluid dynamics code). The simulation is done by varying the several design parameters such as the impeller blade shapes and the gap between the stabilizer and impeller. The computational results are compared with the experimental data at the fan outlet region. Finally some helpful guides for the optimum design of cross-flow fans are proposed.

The aerodynamic performance of a shrouded tail rotor in hover has been studied by using a compressible inviscid flow solver on unstructured meshes. The numerical method is based on a cell­centered finite-volume discretization and an implicit Gauss-Seidel time integration. The results show that the performance of an isolated rotor without shroud compares well with experiment. In the case of a shrouded rotor, correction of the collective pitch angle is made such that the overall performance matches with experiment to account for the uncertainties of the experimental model configuration. Details of the flow field compare well with the experiment confirming the validity of the present method.

The Boltzmann kinetic equation is solved directly by means of the conservative splitting method. Underexpanded supersonic free jet flows with small Knudsen numbers are studied. In this numerical simulation features intrinsic to appropriate experiments are observed. Streamwise vortices in a mixing layer and chaotic downstream temporal-spatial fluctuations of microscopic quantities with large amplitude are obtained.

A mixed analytical/numerical method is developed here to solve the low Reynolds number kepsilon turbulence model. In this method the advection-diffusion part is solved numerically, while the source terms are split into two parts: one part is solved analytically and the next is solved numerically.

Numerical investigations are carried out to study the mass flux and temperature distribution in a calandria using a 3-D RANS code. The computations made for simulations of flow and convective heat transfer with near-to working conditions. The work provides an estimate of the safe working limits of the heat dissipation by virtue of prediction of the 'hot spots' in the calandria. The work assumes significance for preliminary designs of the reactors and for detailed critical parametric analysis that would be otherwise more expensive.

The physical model considered here is a horizontal layer of fluid heated below and cooled above with a periodic array of evenly spaced square cylinders placed at the center of the layer, whose aspect ratio here varies from unity to six. Periodic boundary condition is employed along the horizontal direction to allow for lateral freedom for the convection cells. Two-dimensional solution for unsteady natural convection is obtained using an accurate and efficient Chebyshev spectral multi-domain methodology for a given Rayleigh numbers of $10^6$

This paper deals with a thrust generation of flapping-airfoil by dynamic stall. From many other previous research results, phase angle $ between pitching and plunging mode of flapping motion must be 90 deg. to satisfy maximum propulsive efficiency. In this case, leading edge vortex is relatively small. This phenomenon is related dynamic stall. So preventing leading edge vortex induced by dynamic stall guarantees maximum propulsive efficiency. But, in this paper we insist the leading edge vortex yields quite a positive influence on thrust generation and propulsive efficiency. In order to certify our opinion, pitching and plunging motions were calculated with the parameter of amplitude and frequency by using the unsteady, incompressible Navier-Stokes flow solver with a two-equation turbulence model. For more efficient computation, it is parallelized by MPI programming method.

The complex, unsteady, self-sustained pressure oscillations induced by supersonic flow past a rectangular cavity is investigated using numerical simulations. The present numerical study is performed using a parallel, multiblock solver for the two-dimensional, compressible Navier­Stokes equations. Open cavities with length-to-depth (L / D) ratio in the range 0.5 - 3.3 are considered. This paper sheds light on the cavity physics, cavity oscillatory mechanism, and the organisation of vortical structures inside the cavity. The vortex shedding phenomenon, the shear layer impingement event at the aft wall and the movement of the acoustic/compression wave within the cavity are well predicted. The vortical structures· and the source of the acoustic disturbances are found to be located near the aft wall of the cavity. With the increase in the cavity length, strong recompression of the flow near the aft wall leading to a sudden jump in the cavity form drag is observed. The estimated cavity tones are in good agreement with the available semi­empirical relation. Multiple peaks are noticed in deep and long cavities. For the present free­stream Mach number 1.71, it is observed that around L/D=2.0, the cavity oscillatory mechanism changes from the transverse to longitudinal oscillatory mode. The effects of this transition on various fluid dynamics and acoustic properties are also discussed.