In this paper, some numerical methods recently developed for gas-liquid two-phase flows are reviewed. And then, a preconditioning method to solve cavitating flow by the author is introduced. This method employs a finite-difference Runge-Kutta method combined with MUSCL TVD scheme, and a homogeneous equilibrium cavitation model. So that it permits to treat simply the whole gas-liquid two-phase flow field including wave propagation, large density changes and incompressible flow characteristic at low Mach number. Finally, numerical results such as detailed observations of the unsteady cavity flows, a sheet cavitation break-off phenomena and some data related to performance characteristics of hydrofoils are shown.

An adaptive nonlinear artificial dissipation model is presented for performing aeroacoustic computations by the high-order and high-resolution numerical schemes based on the central finite differences. An effective formalism of it is devised by combining a selective background smoothing term and a well-established nonlinear shock-capturing term which is for the temporal accuracy as well as the numerical stability. A conservative form of the selective background smoothing term is presented to keep accurate phase speeds of the propagating nonlinear waves. The nonlinear shock-capturing term that has been modeled by the second-order derivative term is combined with it to improve the resolution of discontinuities and stabilize the strong nonlinear waves. It is shown that the improved artificial dissipation model with an adaptive control constant which is independent of problem types reproduces the correct profiles and speeds of nonlinear waves, suppresses numerical oscillations near discontinuity and avoids unnecessary damping on the smooth linear acoustic waves. The feasibility and performance of the adaptive nonlinear artificial dissipation model are investigated by the applications to actual computational aeroacoustics problems.

A multigrid DADI method for low Mach number preconditioning method is presented. The eigenvalues of governing equations are modified by A low Mach number preconditioner developed by Choi & Merkle, and it results in an accurate solution and fast convergence In the low Mach number region. The convergence of numerical method is further accelerated by multigrid method. The efficient and accuracy of present method is shown by comparison with conventional solution method for the compressible flows.

The velocity and temperature profiles in the cabin of the automobile affect greatly to the comfort of the passenger. In this paper, the three dimensional flow and temperature analysis in the cabin of the automobile which is geometrically complicated was performed to investigate and predict the velocity and temperature profile. The three dimensional Navier-Stokes code used in this case was validated by performing of a 1/5 experimental scale model vehicle flow anal)rsis successfully. The temperature field of cavity was analyzed for Energy-equation code validation. The comparison of the results are made with the polished computational data and give a coincided one.

The methods of atmospheric boundary layer generation in test section were reviewed. To utilize conventional aerodynamic wind tunnels as atmospheric wind tunnels, boundary layer growth should be accelerated. To achieve this, improvement of boundary layer generation devices is required and it might be done by CFD. In this respect, CFD application cases in boundary generation devices were reviewed and potential areas were considered. Some cases are tried by Fluent 5 code.

A numerical study was made to choose the better turbulence model for the flow in the discharge flow path from a diffuser to a wall. In this study standard $\kappa-\epsilon$ model(SKE), RNG $\kappa-\epsilon$ model(RNG), and Reynolds stress model(RSM) were applied. In case of the flow with relatively high Reynolds number at a diffuser inlet, the pressure loss coefficients by RNG have a tendency to be near to those by SKE at small ratio(below about 0.35) of $h/D_o$, but to those by RSM at large ratio(above about 0.35). At large ratio RNG begins to enlarge the effects of rapid strain and streamline curvature. RNG & RSM are recommended as the appropriate turbulence models for this case. But it is noticeable that the velocity gradient pattern in RNG is same as in SKE, and also that the total pressure distribution in RNG is same as in RSM only at swirling flow area, same as in SKE only at main flow area.

The eddy viscosity turbulence models were applied to predict the flows through a cascade, and the prediction performances of turbulence models were assessed by comparing with the experimental results for a controlled diffusion(CD) compressor blade. The original $\kappa-\omega$ turbulence model and $\kappa-\omega$ shear stress transport(SST) turbulence model were used as two-equation turbulence model which were enhanced for a low Reynolds number flow and the Baldwin-Lomax turbulence model was used as algebraic turbulence model. Farve averaged Wavier-Stokes equations in a two-dimensional, curvilinear coordinate system were solved by an implicit, cell-centered finite-volume computer code. The turbulence quantities are obtained by lagging when the men flow equations have been updated. The numerical analysis was made to the flows of CD compressor blade in a cascade at three different incidence angles (40. 43.4. 46 degrees). We found the reversion in the prediction performance of original $\kappa-\omega$ turbulence model and $\kappa-\omega$ SST turbulence model when the incidence angie increased. And the algebraic Baldwin-Lomax turbulence model showed inferiority to two-equation turbulence models.

Aparallel computing environment was constructed based on Windows 2000 operating system. This cluster was configured using Fast-Ethernet system to hold up together the clients within a network domain. For the parallel computation, MPI implements for Windows such as MPICH.NT.1.2.2 and MP-MPICHNT.1.2 were used with Compaq Visual Fortran compiler which produce a well optimized executives for x86 systems. The evaluation of this cluster performance was carried out using a preconditioned Navier-Stokes code for the 2D analysis of a compressible and viscous flow around a compressor blade. The parallel performance was examined in comparison with those of Linux clusters studied previously by changing a number of processors, problem size and MPI libraries. The result from the test problems presents that parallel performance of the low cost Fast-Ethernet Windows cluster is superior to that of a Linux cluster of similar configuration and is comparable to that of a Myrinet cluster.

The transient flow fields in direct injection engines were analyzed by using the STAR-CD CFD code during the intake/compression processes. The analyses were focused on the computation grid generation by using the IC3M code which is a pre-developed and especially well adapted for the analyses of internal combustion engines. The results showed that the used grid generation technique was well suited for the flow analyses on any internal combustion engines.

This paper presents the results of numerical simulation of wind environment and wind coefficient around super high-rise building. The analysis of aerodynamic response due to wind-induced forces and wind effect to surrounding buildings is important to high-rise building. This paper simulates the wind force to the high-rise building and wind flow pattern around the high-rise building, and shows the usability of CFD analysis to design process of high-rise building. A Navier-Stokes-Solver (FLUENT) with Quick spatial discretization scheme and RNG $\kappa-\epsilon$ turbulence model has been applied to the computation of the three dimensional turbulent flow.

The incompressible Reynolds-averaged Navier-Stokes equations are solved to predict the aerodynamic characteristics of an airship hull. The concept of pseudo-compressibility is employed to couple the pressure field with the velocity field. The upwind differencing method for spatial discretization and a line relaxation scheme for time integration are used. The flowfield around the low drag airship hull of fineness ratio 4 is solved for two Reynolds numbers with a wide range of angle of attack. The effect of Reynolds number and transition position is briefly examined together with the change in aerodynamic coefficients due to a gondola attached to the hull, and the results will be used as basic data for the design of a low drag airship hull.

Secondary instability of flow past a circular cylinder is examined using direct numerical simulation at Reynolds number 220 and 250. The higher-order finite difference scheme is employed for the spatial distributions along with the second order Adams-Bashforth and the first order backward-Euler time integration. In x-y plane, the convection term is applied by the 5th order upwind scheme, and the pressure and viscosity terms are applied by the 4th order central difference. In spanwise, Navier-Stokes equation is distributed using Spectral Method. The critical Reynolds number for this instability is found to be about Re=190. The secondary instability leads re three-dimensionality with a spanwise wavelength about 4 cylinder diameters at onset (A-mode). Results of three-dimensional effect in wake of a circular cylinder are represented with spanwise and streamwise vorticity contours as Reynolds numbers.

The thin-layer Navier-Stokes equations are solved for the hypersonic flow over blunt cone configurations with applications to laminar as well as turbulent flows. The equations are expressed in the forms of flux-vector splitting and explicit algorithm. The upwind schemes of Steger-Warming and van Leer are investigated in their ability to accurately predict the heating loads along the surface of the body. A comparison with the second order extensions of these schemes is made and a hybrid scheme incorporating a combination of central differencing and flux-vector-splitting is presented. This scheme is also investigated in its ability to accurately predict heat transfer distributions.

Flow computations have been conducted to study the impingement flowfield over the KSR-III flume deflector To validate Euler solver for the jet impingement flowfieid, the jet flow over a double wedge deflector have been calculated and showed reasonable agreement with experimental data. The transient flow behavior of flume over deflector have been investigated and the flume from the rocket nozzle proved to be getting out of the deflector safely and the thermal effect on the base region of rocket was not considerable.

This paper presents the numerical methodology of ATHOS3 code for thermal hydraulic analysis of Pressurized Water Reactor (PWR) steam generators. Topics include porous media approach, governing equations, physical models and correlations for solid-to-fluid interaction and heat transfer, and numerical solution scheme. The ATHOS3 code is applied to the thermal hydraulic analysis of steam generator in the Korea Kori Unit-1 nuclear power plant and the computed results are presented.

A numerical investigation is made of unsteady double-diffusive convection of a Boussinesq fluid in a rectangular cavity subject to time-periodic thermal excitations. The fluid is initially stratified between the top endwall of low solute concentration and the bottom endwall of high solute concentration. A time-dependent heat flux varying in a square wave fashion, is applied on one sidewall to induce buoyant convection. The influences of the imposed periodicity on double-diffusive convection are scrutinized. A special concern is on the occurrence of resonance that the fluctuations of flow and attendant heat and mass transfers are mostly amplified at certain eigenmodes of the fluid system. Numerical solutions are analyzed to illustrate the characteristic features of resonant convection.

The performance prediction of SMART MCP was performed using a computational fluid dynamics code. General capacity-head performance curve of MCP, which is provided to other design branches as design input, was obtained and it showed the typical type of axial pump performance curve. When four MCPs operate in parallel and one of them stops while the others continue to operate, SMART requires reduced power operation. A procedure for predicting the performance of SMART MCP for that case was developed and verified with available experimental data. An analysis based on the developed procedure was performed for two cases; the impeller of sloped MCP is fixed or free to rotate in reverse direction. According to the results, $73\%$ flow rate of normal operation enters the reactor core in the case of the locked impeller. In case of the impeller free rotation, the flow rate entering the reactor core is $62.8\%$.

충격파-와동 간섭은 다양한 비선형 물리 현상을 포함한다. 그 중에서도, 충격파 분리와 충격파-충격파 간섭에 대해서는 잘 알려져 있는 반면, 미끄럼 층에서 반사되는 파동이나 관통 충격파 등에 대해서는 지금까지 적절한 연구가 이루어지지 못했다. 저자들은 복잡한 충격파-와동 간섭이 시간에 따라 몇 개의 국지적인 소모델 (submodels)로 귀착됨을 발견하였다. 이는 와동에 의한 충격파의 분리, 충격파-충격파 간섭, 충격파-미끄럼층 간섭, 와동 중심의 충격파통과, 그리고 충격파-소와동 간섭이다. 이러한 5 개의 소모델은 탐구 범위 내에서 충격파-와동 간섭의 전체 구조를 구성한다.

We used In-line orifice mixer for efficient chemicals mixing in water treatment. The method of using In-line orifice mixer has been already proved the improvement of water treatment efficiency. Code of computational fluid dynamics for numerical analysis was performed using FLUENT, a commercial code. As variable for exactly standardizing, a proper ratio between an outer diameter of deflector and a diameter of pipe, the distance between deflector and orifice, a determination of orifice diameter fur an optimal mixing, a distance between injection nozzle's position and cone, Numerical study has been performed for optimal standard and analyzed flow field on a basis of turbulent intensity in an orifice downstream.

An experimental model to enhance the mixing of parallel supersonic-subsonic jet ($M_1$=1.78 and $M_2$=0.30) is simulated with a numerical technique by modeling the wall-mounted cavity to a boundary condition of oscillating pressure. The computed pilot pressure distributions along three representative cross sections show a good agreement with the equivalent experimental data. The irradiation of acoustic wave in the ultrasonic range causes the mixing augmentation of jet and wake due to the transfer of vibration energy between fluid particles.

NURBS 곡면식으로 정의된 물체 표면상에서 표면 정렬 격자를 생성할 수 있는 방법을 소개하였다. 공학 응용분야에서의 물체 표면 정의는 여러 개의 패치들로 표현되는 것이 일반적이고, 여기서 소개하는 표면격자 생성기법은 이러한 여러 패치들에 걸쳐서 분포되는 정렬격자를 쉽게 생성할 수 있도록 한다 이 기법은 매개변수 형태의 타원형 격자생성 방정식의 해를 구하되, 여러 NURBS 패치에 걸쳐서 투영/분포된 초기 격자계를 타원형 방정식 반복계산 과정의 매개변수형 표면 정의식으로 임시 활용한다. 매개변수형 타원형 방정식의 해가 얻어지고 나면, 그 결과 격자계를 다시 NURBS 패치에 투영을 시키고 타원형 방정식의 해를 구하는 과정이 반복된다. 이러한 반복과정이 전체적으로 수렴이 이루어질 때까지 반복된다. 이 방법에 의해서 얻어지는 표면 정렬 격자계들은 타원형 격자생성기법의 특징인 완만성을 가지면서 정의된 물체표면에서 벗어나지 않는 격자점들이 된다. 소개된 방법은 간단하면서도 하나의 NURBS 곡면만이 아니라 여러 개의 NURBS 곡면에 걸쳐있는 정렬격자계를 효율적으로 생성할 수 있도록 해주며, 그 기본적인 접근법은 NURBS 곡면식 만이 아니라 다른 형태의 매개변수형 형상 정의식에도 적용이 가능하다.

Three dimensional orthogonal grid generation is able to control effectively the grid spacing near the boundaries, but there are some difficulty to meshing complex geometry. The mesh complex geometry by orthogonal grid generation method must divide block of geometry It is required a careful skill, and long time. Its also difficulty to make unstructured mesh on complex geometry. Particularly, three dimensional geometry must have more time and effort. Recently, there have been growing interests in mesh generation of complex grometry, aslike an automobile headlamp, the heart. The method of easily meshing complex geometry is resarched to solve them. We suggest octree grid into one among these methods. As octrce grid is automaticaly adapted at the boundaries by determine the level operations to control the grid spacing near the boundaries are unnecessary. In this paper we showed throe dimensional mesh generation, and heat-flow analysis on the octree mesh.

In the present study a mesh generation program is developed for three dimensional flow analyses with complex geometry. By the present program one can define vertices using various coordinate systems and cells for finite volume approach. Rendered display of the generated mesh can be also available in both orthographic and perspective projection modes. Through perspective projection mode, one can check the quality of generated mesh by moving around inside the mesh like a virtual reality. The examples of the program execution is given in the paper.

Advanced countries centered by National Supercomputing Center are inclined to construct Grid Infra and develop key applications in high performance computing field. They are also trying to globalize the Grid project aided by research groups in nations and continents. Computing technology and application development in Grid computing environment become indirect capital of high performance computing and information technology. Therefore, Korean government would like to participate in their Grid construction and application development actively and pursue to Grid project to develop Grid industries such as IT, BT. next five years.

The Gird is a communication service that collaborates dispersed high performance computers, large-scale databases and modern equipments so that those can be shared and worked together. In this paper. CFD research based on National Grid project is discussed. To validate the Grid technology, the flow past ONERA M6 wing and the flow past infinite wing are simulated on the National Grid testbed.

A project has been initiated at KISTI Supercomputing Center at year 1999 to build a Tflops cluster in order to support newly emerging demands for larger scale parallel computing resources. As a supportive substitute of traditional supercompters, the cluster systems are expected to play important role in supporting diversified computational needs. In this paper, KISTI Supercomputing Center's working status of building Tflops cluster system and future plans.