• Advances in aircraft and spacecraft science
• /
• v.9 no.5
• /
• pp.415-431
• /
• 2022

Secondary flows have a huge impact on losses generation in modern low pressure gas turbines (LPTs). At design point, the interaction of the blade profile with the end-wall boundary layer is responsible for up to 40% of total losses. Therefore, predicting accurately the end-wall flow field in a LPT is extremely important in the industrial design phase. Since the inlet boundary layer profile is one of the factors which most affects the evolution of secondary flows, the first main objective of the present work is to investigate the impact of two different inlet conditions on the end-wall flow field of the T106A, a well known LPT cascade. The first condition, labeled in the paper as C1, is represented by uniform conditions at the inlet plane and the second, C2, by a flow characterized by a defined inlet boundary layer profile. The code used for the simulations is based on the Discontinuous Galerkin (DG) formulation and solves the Reynolds-averaged Navier-Stokes (RANS) equations coupled with the Spalart Allmaras turbulence model. Secondly, this work aims at estimating the influence of viscosity and turbulence on the T106A end-wall flow field. In order to do so, RANS results are compared with those obtained from an inviscid simulation with a prescribed inlet total pressure profile, which mimics a boundary layer. A comparison between C1 and C2 results highlights an influence of secondary flows on the flow field up to a significant distance from the end-wall. In particular, the C2 end-wall flow field appears to be characterized by greater over turning and under turning angles and higher total pressure losses. Furthermore, the C2 simulated flow field shows good agreement with experimental and numerical data available in literature. The C2 and inviscid Euler computed flow fields, although globally comparable, present evident differences. The cascade passage simulated with inviscid flow is mainly dominated by a single large and homogeneous vortex structure, less stretched in the spanwise direction and closer to the end-wall than vortical structures computed by compressible flow simulation. It is reasonable, then, asserting that for the chosen test case a great part of the secondary flows details is strongly dependent on viscous phenomena and turbulence.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.30 no.1
• /
• pp.47-53
• /
• 2017

A boundary layer integral combined with a 1-D isentropic core flow model has been successfully used to determine heat transfer rate on the surface of a supersonic nozzle. However its accuracy is affected by the core flow condition which is used as a boundary condition for the integral calculation. Because flow behavior near a nozzle throat deviates from 1-D isentropic condition due to 2-D flow turning and interaction between core flow and boundary layer, accuracy of heat transfer calculation decreases at a nozzle throat. Therefore, CFD is adopted to deduce improved core flow condition and increase accuracy of boundary layer integral at nozzle throat in this research. Euler model and SST $k-{\omega}$ model is solved by CFD code and used as a boundary condition for boundary layer integral. Developed code is tested in the supersonic nozzle from the previous research and improvement in accuracy is observed, especially at nozzle throat and diverging section of the nozzle. Error between experimental result and calculation result reduced by 16% when a calculation is made based on the SST $k-{\omega}$ model. Method developed in this research is expected to be used in thermal design of the rocket nozzle.

• Proceedings of the KSME Conference
• /
• 2001.06e
• /
• pp.355-360
• /
• 2001

This study addresses a computational work of the impulsive wave which is discharged from the open end of a pipe. An initial compression wave inside the pipe is assumed to propagate toward atmosphere. The over pressure and wave-length of the initial compression wave are changed to investigate the characteristic values of the impulsive wave. The second order total variation diminishing (TVD) scheme is employed to solve the axisymmetric, compressible, unsteady Euler equations. The relationship between the initial compression wave form and impulsive wave is characterized in terms of the peak pressure of the impulsive wave and its directivity. The results obtained show that for the initial compression wave of a large wave-length the peak pressure of the impulsive wave does not depend on the over pressure of the initial compression wave, but for the initial compression wave of a very short wave-length, like a shock wave, the peak pressure of the impulsive wave is increased with an increase in the over pressure of the initial compression wave. The directivity of the impulsive wave to the pipe axis becomes significant with a decrease in the wave-length of the initial compression wave.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.793-798
• /
• 2004

The vibration of a curved pipe conveying fluid is studied when the pipe is clamped at both ends. To consider the geometric nonlinearity, this study adopts the Lagrange strain theory for large deformation and the extensible dynamics based on the Euler-Bernoulli beam theory for slenderness assumption. By using the extended Hamilton principle, the non-linear partial differential equations are derived for the in-plane motions of the pipe. The linear and non-linear terms in the governing equations are compared with those in the previous study, and some significant differences are discussed. To investigate the vibration characteristics of the system, the discretized equations of motion are derived from the Galerkin method. The natural frequencies varying with the flow velocity are computed from the two cases, which one is the linear problem and the other is the linearized problem in the neighborhood of the equilibrium position. From these results, we should consider the geometric nonlinearity to analyze the dynamics of a curved pipe conveying fluid more precisely.

• 한국전산유체공학회:학술대회논문집
• /
• 2001.10a
• /
• pp.84-90
• /
• 2001

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.

• 한국전산유체공학회:학술대회논문집
• /
• 2008.03a
• /
• pp.208-214
• /
• 2008

The equilibrium molecular-dynamic simulations have been performed to estimate the properties of the three kinds of fluids (the Lennard-Jones fluid, water and aqueous sodium-chloride solution) confined between two plates that are separated by 1.086 nm; included in the equilibrium properties are the density distribution and the static structure, and the diffusivity in the dynamic property. Three kinds of fluids considered in this study are. The water molecules are modeled by using the SPC/E model and the ions by the charged Lennard-Jones particle model. To treat the water molecules, we combined the quaternion coordinates with Euler angles. We also proposed a plausible algorithm to assign the initial position and direction of molecules. The influence of polarization of water molecules as well as the presence of ions in the solution on the properties will be addressed in this study. In addition, we performed the non-equilibrium molecular-dynamic simulation to compute the flow velocity for the case with the gravitational force acting on molecules.

• Proceedings of the KSME Conference
• /
• 2003.04a
• /
• pp.1572-1577
• /
• 2003

High-power pulsed laser ablation under atmospheric pressure is studied utilizing numerical and experimental methods with emphasis on recondensation ratio, and the dynamics of the laser induced vapor flow. In the numerical calculation, the temperature pressure, density and vaporization flux on a solid substrate are first obtained by a heat-transfer computation code based on the enthalpy method, and then the plume dynamics is calculated by using a commercial CFD package. To confirm the computation results, the probe beam deflection technique was utilized for measuring the propagation of a laser induced shock wave. Discontinuities of properties and velocity over the Knudsen layer were investigated. Related with the analysis of the jump condition, the effect of the recondesation ratio on the plume dynamics was examined by comparing the pressure, density, and mass fraction of ablated aluminum vapor. To consider the effect of mass transfer between the ablation plume and air, unlike the most previous investigations, the equation of species conservation is simultaneously solved with the Euler equations. Therefore the numerical model computes not only the propagation of the shock front but also the distribution of the aluminum vapor. To our knowledge, this is the first work that employed a commercial CFD code in the calculation of pulsed ablation phenomena.

• 한국전산유체공학회:학술대회논문집
• /
• 2006.10a
• /
• pp.31-35
• /
• 2006

The delta-formulation of the Navier-Stokes equations has been popularly used in the aerodynamics area. Implicit algorithm can be easily implemented in that by using Taylor series expansion. This formulation is extended for an unsteady analysis by using a dual-time integration. In the meanwhile, the incompressible flows with heat transfers which occur in the area of thermo-hydraulics have been solved by a segregated algorithm such as the SIMPLE method, where each equation is discretised by using an under-relaxed deferred correction method and solved sequentially. In this study, the dual-time delta formulation is implemented in the segregated Navier-Stokes solver which is based on the collocated cell-centerd scheme with un unstructured mesh FVM. The pressure correction equation is derived by the SIMPLE method. From this study, it was found that the Euler dual-time method in the delta formulation can be combined with the SIMPLE method.

• 한국전산유체공학회:학술대회논문집
• /
• 2008.10a
• /
• pp.208-214
• /
• 2008

The equilibrium molecular-dynamic simulations have been performed to estimate the properties of the three kinds of fluids (the Lennard-Jones fluid, water and aqueous sodium-chloride solution) confined between two plates that are separated by 1.086 nm; included in the equilibrium properties are the density distribution and the static structure, and the diffusivity in the dynamic property. Three kinds of fluids considered in this study are. The water molecules are modeled by using the SPC/E model and the ions by the charged Lennard-Jones particle model. To treat the water molecules, we combined the quaternion coordinates with Euler angles. We also proposed a plausible algorithm to assign the initial position and direction of molecules. The influence of polarization of water molecules as well as the presence of ions in the solution on the properties will be addressed in this study. In addition, we performed the non-equilibrium molecular-dynamic simulation to compute the flow velocity for the case with the gravitational force acting on molecules.

• Proceedings of the KSME Conference
• /
• 2001.11b
• /
• pp.640-645
• /
• 2001

The present study is to investigate the characteristics of the impulse waves discharged from the exit of the convergent and divergent pipes. An experiment is carried out using a shock tube with an open end and is compared to the computation of the axisymmetric, compressible, unsteady Euler equations, which are solved by the second-order total variation diminishing(TVD) scheme. For the computational work, some initial compression waves are assumed inside the pipe so that those are identical to the experimental ones of the shock tube. The results show that the peak pressures of the impulse waves discharged from the exit of convergent and divergent pipes decrease with an increase in the wavelength of the initial compression wave. All of the impulse waves have a strong directivity toward the pipe axis, regardless of the exit type of the pipe employed. The impulse waves discharged from the divergent pipe are stronger than those from the straight pipe, while the impulse waves of the convergent pipe are weaker than those from the straight pipe. It is believed that the convergent pipe can playa role of a passive control to reduce the peak pressure of the impulse wave. The present computations represent the experimented impulse waves with a good accuracy.