• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2022년도 학술발표회
• /
• pp.247-247
• /
• 2022

제트류는 복잡한 흐름 중 하나로 다양한 크기의 에디가 다양한 운동량을 가지고 있다. 이러한 제트류를 구현하기 위해서는 난류 운동 에너지 등 제트류의 특성을 잘 반영하여야 한다. 제트를 구현하기 위해서는 수리학적 모델, 현장 실험 등 많은 방법이 있으며, 본 연구에서는 상대적으로 공간, 시간적 비용이 적게 드는 수치해석 방법을 사용하여 연구를 진행하였다. 대표적인 수치해석방법에는 DNS(Direct Numerical Simulation), LES(Large Eddy Simulation), RANS(Reynolds Averaged Navier Stokes) 등이 있다. RANS는 시간 평균 흐름 특성만 산출하며 제트의 복잡성을 재현하는 데 한계가 있어, 본 연구는 DNS와 LES 모델을 이용하여 제트류를 구현하는 것에 초점을 맞추었다. DNS는 해당 격자에서 발생하는 모든 에디를 직접 해석 때문에 난류 모델링이 필요하지 않지만, 많은 수의 그리드가 필요하여 수치해석 시 소요시간이 긴 편이다. LES는 대규모 에디는 직접 해석하지만 일정 크기 이하의 소용돌이를 해석하기 위해서 모델이 필요하다. 따라서 서브 그리드 모델에 따라 약간 다른 결과를 보인다. 이러한 문제점을 해결하기 위해 본 연구에서는 LES의 기존 서브 그리드 모델을 사용하지 않고 신경망 모델로 학습한 DNS 결과를 활용하는 방법을 제안한다. 우선 DNS와 LES 모델을 사용하여 에너지 스펙트럼을 비교하여 서브 그리드 모델이 시작하는 파수를 찾는다. 이후 특정 파수 아래의 작은 에디를 모사할 적절한 신경망 모델을 결정하여 DNS의 작은 에디를 신경망 알고리즘이 모사할 수 있도록 학습시킨다. 이후 기존 서브 그리드 모델을 사용하지 않고 학습된 신경망 알고리즘을 사용한 LES 모델이 모사한 제트류와 실제 DNS 모델을 사용한 제트류를 비교 및 평가한다.

• 대한기계학회논문집B
• /
• 제31권9호
• /
• pp.799-806
• /
• 2007

Direct Numerical Simulation was carried out to predict mass transfer in turbulent flow around a rotating stepped cylinder. This investigation is a follow-up study of DNS of turbulent flow in Nesic et al. [Corrosion, Vol. 56, No. 10, pp. 1005 - 1014] The original motivation of this work stemmed from the efforts to design a simple device which can generate flows of high turbulence intensity at low cost for corrosion researchers. Two cases were considered; Sc=1 and 10 both at Re=335. Here, Sc and Re stand for Schmidt number and Reynolds number, respectively, based on the step height and the surface speed of the cylinder upstream of the step. Main focus was placed on the correlation between turbulence and mass transfer. The spatio-temporal evolution of concentration field is discussed. The numerical results are qualitatively compared with those of the experiment conducted with a similar flow configuration.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2002년도 학술대회지
• /
• pp.827-830
• /
• 2002

A direct numerical simulation is performed for turbulent heat transfer in a concentric annulus at $Re_{Dh}=8900\;and\;Pr=0.71$ for two radius ratios ($R_{1}/R_{2}=0.1\;and\;0.5$) and $q^{\ast}=1.0$. Main emphasis is placed on the transverse curvature effect on near-wall turbulent thermal structures. Near-wall turbulent structures close to the inner and outer walls are scrutinized by computing the lower-order statistics. The fluctuating temperature variance and turbulent heat flux budgets are illustrated to confirm the results of the lower-order statistics. The present numerical results show that the turbulent structures near the outer wall are more activated than those near the inner wall, which may be attributed to the different vortex regeneration processes between the inner and outer walls.

• 한국전산유체공학회지
• /
• 제14권4호
• /
• pp.101-108
• /
• 2009

We present a new finite-element scheme for direct numerical simulation of particle suspensions in simple shear flow of a viscoelastic fluid in 3D. The sliding tri-periodic representative cell concept has been combined with DEVSS/DG finite element scheme by introducing constraint equations along the domain boundary. Rigid body motion of the freely suspended particle is described by the rigid-shell description and implemented by Lagrangian multipliers on particle boundaries. We present the bulk rheology of suspensions through the numerical examples of single-, two- and many-particle problems, which represent a large number of such systems in simple shear flow. We report the steady bulk viscosity and the first normal stress coefficient, which show shear-thickening behavior for both properties.

• Journal of Ship and Ocean Technology
• /
• 제9권4호
• /
• pp.1-10
• /
• 2005

The effect of buoyancy orientation on turbulent channel flow has been investigated using DNS (direct numerical simulation). Grashof number is kept at 9.6 $\times 10^{5}$ while changing the orientation of the buoyancy vector to be parallel or perpendicular to the channel walls. Four study cases can be distinguished during this research namely; streamwise, wall-normal unstable stratification, wall-normal stable stratification and spanwise oriented buoyancy. The driving mean pressure gradient used in all cases is adjusted to keep mass flow rate constant while friction Reynolds number is around 150. At this Grashof number, the skin friction shows decrement in the unstable and stable stratification and increment in the other two cases. Analyses of the changes of flow structure for the four cases are presented highlighting on the mean quantities and second order statistics.

• 대한기계학회논문집B
• /
• 제31권9호
• /
• pp.780-789
• /
• 2007

Lock-on phenomenon in the wake of a circular cylinder is investigated at the Reynolds number of 360 using direct numerical simulation (DNS). To induce lock-on, a streamwise velocity perturbation with a frequency of twice the natural shedding frequency is superimposed on the free stream velocity. The Reynolds stress distributions are investigated to analyze the streamwise force balance acting on the recirculation region and the results are compared with the previous experimental result. When the lock-on occurs, the pressure force on the recirculation region is shown to increase mainly due to the reversal of the Reynolds shear stress distribution, which is consistent with our previous results using PIV measurement. It is also shown that, with the lock-on, the strength of the primary vortices increases whereas that of the secondary vortices decreases significantly. Further, under the lock-on condition the wavelength of the secondary vortices increases by as much as 2.5 times.

• Nuclear Engineering and Technology
• /
• 제37권6호
• /
• pp.511-524
• /
• 2005

Thermalhydraulic reactor simulation of tomorrow will require a new generation of codes combining at least three scales, the CFD scale in open medium, the component scale and the system scale. DNS will be used as a support for modelling more macroscopic models. NEPTUNE is such a new generation multi-scale platform developed jointly by CEA-DEN and EDF-R&D and also supported by IRSN and FRAMATOME-ANP. The major steps towards the next generation lie in new physical models and improved numerical methods. This paper presents the advances obtained so far in physical modelling for each scale. Macroscopic models of system and component scales include multi-field modelling, transport of interfacial area, and turbulence modelling. Two-phase CFD or CMFD was first applied to boiling bubbly flow for departure from nucleate boiling investigations and to stratified flow for pressurised thermal shock investigations. The main challenges of the project are presented, some selected results are shown for each scale, and the perspectives for future are also drawn. Direct Numerical Simulation tools with Interface Tracking Techniques are also developed for even smaller scale investigations leading to a better understanding of basic physical processes and allowing the development of closure relations for macroscopic and CFD models.

• 한국해양공학회지
• /
• 제33권5호
• /
• pp.387-393
• /
• 2019

In this study, the free motion of a riser due to vortex shedding was numerically simulated with Large Eddy Simulation (LES) and Detached Eddy Simulation (DES) turbulence models. A numerical simulation program was developed by applying the Rhie-Chow interpolation method to the pressure correction of the OpenFOAM standard solver pimpleDyMFoam. To verify the developed program, the vortex shedding around the fixed riser at Re = 3900 was calculated, and the results were compared with the existing experimental and numerical data. Moreover, the vortex-induced vibration of a riser supported by a linear spring was numerically simulated while varying the spring constant. The results are compared with published direct numerical simulation (DNS) results. The present calculation results show that the numerical method is appropriate for simulating the vortex-induced motion of a riser, including lock-in phenomena.

• 한국안전학회지
• /
• 제25권1호
• /
• pp.9-16
• /
• 2010

Flame structure of diffusion flame interacting with a single vortex was investigated with direct numerical simulation (DNS). A well-known counterflow diffusion flame was used as an initial flat flame and single vortices were made by issuing a high-velocity jet abruptly in fuel- and air-side. The variations in the maximum concentration of major species (CO and $CO_2$) and NOx (NO and $NO_2$) with the stoichiometric scalar dissipation rate were investigated. Unsteady effects in the species concentration variation of the flame interacting with a vortex were identified by comparing with that of steady flame. $NO_2$ formation characteristics of the flame interacting with a vortex were well understood by investigating the $HO_2$ formation. To enhance the prediction performance in the fire simulation, current turbulent combustion modelings are needed to be modified by adopting the unsteady effects in the species concentrations of diffusion flame interacting with a vortex.

• 대한기계학회논문집B
• /
• 제29권3호
• /
• pp.314-321
• /
• 2005

Discrete Wavelet Transform (DWT) has been applied to the Direct Numerical Simulation (DNS) data of turbulent channel flow. DWT splits the turbulent flow into two orthogonal parts, one corresponding to coherent structures and the other to incoherent background flow. The coherent structure is extracted from not vorticity field but velocity's since the channel flow is not isoropic. By comparing DWT's result of channel flow with that of isotropic flow, it is shown that coherent structure maintains the properties of original channel flow. The velocity field of coherent structures can be represented by few wavelet modes and that these modes are sufficient to reproduce the velocity probability density function (PDF) and the energy spectrum over the entire inertial range. The remaining incoherent background flow is homogeneous, has small amplitude, and is uncorrelated. These results are compared with those obtained for the same compression rate using large eddy simulation (LES) filtering. In contrast to the incoherent background flow of DWT, the LES subgrid scales have a much larger amplitude and are correlated, which makes their statistical modeling more difficult.