• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.40 no.5
• /
• pp.375-384
• /
• 2012

This paper presents numerical results comparing the performance of the 2008 Wilcox $\mathcal{k}-{\omega}$ turbulence model to the one of the 1988 Wilcox $\mathcal{k}-{\omega}$ model for supersonic flows. A comparison with experimental data is offered for a shock wave/turbulent boundary layer interaction case and two ramp injector mixing cases. Furthermore, a comparison is performed with empirical correlations on the basis of skin friction for flow over a flat plate and shear layer growth for a free shear layer. It is found that the maximum injectant mass fraction of some ramp injector cases is better predicted using the 1988 Wilcox model. On the other hand, the 2008 model performs better in simulating shock-boundary layer cases.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.5 no.3
• /
• pp.25-32
• /
• 2001

A passive control method of the interaction between a weak normal shock-wave and a turbulent boundary-layer was simulated using two-dimensional Navier-Stokes computations. The inflow Mach number just upstream of the normal shock wave was 1.33. A porous plate wall having a cavity underneath was used to control the shock-wave/turbulent boundary-layer interaction. The flows through the porous holes and inside the cavity were investigated to get a better understanding of the flow physics involved in this kind of passive control method. The present computations were validated by some recent wind tunnel tests. The results showed that downstream of the rear leg of the $\lambda$-shock wave the main stream inflows into the cavity, but upstream of the rear leg of the $\lambda$-shock wave the flow proceeds from the cavity toward to the main stream. The flow through the porous holes did not choke fur the present shock/boundary layer interaction.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.32 no.10
• /
• pp.1-11
• /
• 2004

An experimental study was carried out to investigate the influence of flow conditions of a boundary layer on the near-wake of a flat plate. Various attaching positions of tripping wires were selected to change flow conditions on a boundary layer. Laminar, transitional, and turbulent boundary layer conditions at 0.98C from the leading edge are imposed to investigate the evolution of symmetric and asymmetric wake. An x-type hot-wire probe(55P61) is employed to measure at 8 stations of the near-wake region. Measured mean velocity distributions are presented in terms of similarity parameter. It is found that the symmetric wake collapses well to the universal profile in the central part of the wake. However, the universal profile is not suitable in describing an asymmetric wake.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.33 no.12
• /
• pp.977-982
• /
• 2009

Direct numerical simulation (DNS) of a turbulent boundary layer with moderate Reynolds number was performed to scrutinize streamwise-coherence of hairpin packet motions. The Reynolds number based on the momentum thickness (${\theta}_{in}$) and free-stream velocity (U${\infty}$) was varied in the range $Re_{\theta}$=1410${\sim}$2540 which was higher than the previous numerical simulations in the turbulent boundary layer. In order to include the groups of hairpin packets existing in the outer layer, large computational domain was used (more than 50${\delta}_o$, where ${\theta}_o$ is the boundary layer thickness at the inlet in the streamwise domain). Characteristics of packet motions were investigated by using instantaneous flow fields, two-point correlation and conditional average flow fields in xy-plane. The present results showed that a train of hairpin packet motions was propagating coherently along the downstream and these structures induced the very large-scale motions in the turbulent boundary layer.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.32 no.6
• /
• pp.463-470
• /
• 2008

Turbulent coherent structures near rod-roughened wall are investigated by analyzing the database of direct numerical simulation of turbulent boundary layer. The surface roughness rods with the height $k/{\delta}=0.05$ are arranged periodically in $Re_{\delta}=9000$. The roughness sublayer is defined as two-point correlations are not independent of streamwise locations around roughness. The roughness sublayer based on the two-point spatial correlation is different from that given by one-point statistics. Quadrant analysis and probability-weighted Reynolds shear stress indicate that turbulent structures are not affected by surface roughness above the roughness sublayer defined by the spatial correlations. The conditionally-averaged flow fields associated with Reynolds shear stress producing Q2/Q4 events show that though turbulent vortices are affected in the roughness sublayer, these are very similar at different streamwise locations above the roughness sublayer. The Reynolds stress producing turbulent vortices in the log layer ($y/{\delta}=0.15$)have almost the same geometrical shape as those in the smooth wall-bounded turbulent flows. This suggests that the mechanism by which the Reynolds stress is produced in the log layer has not been significantly affected by the present surface roughness.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.244-244
• /
• 2021

난류흐름 거동은 지형이나 수공구조물과 같은 고체 경계면의 변화에 민감하게 반응하며 특징 또한 다양하다. 보나 여수로 등과 같은 단차 구조물을 통과하는 흐름은 구조물의 모서리 같은 흐름 경계면이 급변하는 지점에서는 흐름분리(flow separation)가 발생하는 것이 특징이다. 이러한 흐름분리로 인해 전단층이 발생하며 흐름 재순환(recirculation)이 구조물 하류부에 형성된다. 이 연구에서는 낙차공 형식의 단차 구조물 하류부에서의 흐름 거동을 이해하기 위해 CFD모델링을 통하여 계산된 3차원 유동장을 분석한다. 난류 모의는 하이브리드 LES(large-eddy simulation)/RANS 계산 기법인 IDDES(improved delayed detached-eddy simulation)기법을 적용한다. IDDES의 기본 모형으로는 k-ω SST모형과 Spalart-Allmaras모형을 이용하여 두 모형의 성능을 평가한다. 자유수면의 변동은 VoF(volume of fluid)기법을 이용하여 계산하며, 각 지배방정식은 최소의 수치분산을 유지하면서 수치해의 안정성을 확보할 수 있는 2차 정확도의 유한체적법을 이용하여 이산화하였다. 수치해석 결과는 레이놀즈수 23,400과 후르드수 0.22의 조건에서 기존에 계측된 자료와 비교하여 수치모형의 정확도를 평가하고 하상 단차 하류부에서의 흐름 거동 특성을 분석한다. 계산 결과는 공학적으로 널리 사용되는 RANS 수치모의에서 볼 수 없는 전단층과 난류구조의 동적 거동 특성과 이에 따른 레이놀즈 응력분포의 특성을 설명해준다.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.24 no.1
• /
• pp.92-100
• /
• 2018

Vortical systems are considered a main feature to sustain turbulence in a boundary layer through interaction. Such turbulent structures result in frictional drag and erosion or vibration in engineering applications. Research for controlling turbulent flow has been actively carried out, but in order to show the effect of vortices in a turbulent boundary layer, it is necessary to clarify the mechanism by which turbulent energy is transferred. For this purpose, it is convenient to demonstrate and capture phenomena in a laminar boundary layer. Therefore, in this study, the interactions of disturbed flow around a hemisphere on a flat plate in laminar flow were analyzed. In other words, a street of hairpin vortices was generated following a wake region formed after flow separation occurred over a hemisphere. Necklace vortices surrounding the hemisphere also appeared due to a strong adverse pressure gradient that brought high momentum fluid into the wake region thereby leading to an increase in the frequency of hairpin vortices. To mitigate the effect of these necklace vortices, local suction control was applied through a hole in front of the hemisphere. Flow visualization was recorded to qualitatively determine flow modifications, and hot-film measurements quantitatively supported conclusions on how much the power of the hairpin vortices was reduced by local wall suction.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.32 no.7
• /
• pp.518-528
• /
• 2008

The effects of surface roughness on a turbulent boundary layer (TBL) were investigated using particle image velocimetry (PIV). The roughness elements used were periodically arranged two-dimensional spanwise rods, and the roughness height was ${\kappa}/{\delta}$. Introduction of the roughness elements increased the wake strength and the turbulent stress not only in the roughness sublayer but also in the outer layer. This indicates the existence of interaction between inner and outer layers for 2D rod-roughened wall. Roughness effects on a turbulence structure near the wall were obtained by PIV measurements. Iso-contours of mean velocities and Reynolds stresses in the roughness sublayer showed a very good agreement with previous DNS results.

• Journal of KSNVE
• /
• v.11 no.2
• /
• pp.226-233
• /
• 2001

It is essential to analyze structural vibrations due to turbulent wall pressure fluctuations over a body surface which moves through a fluid, because the vibrations can be a severe source of noise affecting to passengers in airplanes and SONAR performance. Generally, this kind of problems have been solved for very simplified models, e.g. plates, which can be applied to the wavenumber domain analysis. In this paper, a finite element modeling of the walt pressure fluctuations is investigated, which can be applied to those over arbitrary smooth surfaces. It is found that the modeled wall pressure fluctuation at nodes becomes uncorrelated at higher frequencies and at lower flow speeds, and the response is over-estimated due to the aliased power. Then the frequency range available for uncorrelated loading model and two power correction schemes are presented.

• Journal of KSNVE
• /
• v.6 no.5
• /
• pp.671-677
• /
• 1996

This paper presents the analytical method for predicting turbulence- induced noise in the multilayered cylinder composed of an outer hose, an inner fluid and an internal core. It is assumed that an infinite axisymmetric cylinder is located horizontally in water with free stream velocity and the turbulent boundary layer (TBL) surrounding the outer hose is fully developed and homogeneous. The transfer function at the core surface due to the propagation of the pressure fluctuation within the TBL is formulated using the linearized Navier-Stockes equation for solid and fluid. In the estimation of the energy spectrum of wall pressure fluctuation, the empirical formula proposed by Strawderman based on the Corcos model is used. A general algorithm for the calculation of the pressure level at the surface of a core, that is, turbulence- induced noise, is presented. Through the detailed numerical simulation, it is found that the major noise mechanism is the propagation of the bulge wave along hose.