Prediction of Unsteady Loading on a Circular Cylinder in High Reynolds Number Flows Using Large Eddy Simulation

Author(s):  
Sung-Eun Kim ◽  
L. Srinivasa Mohan

Large eddy simulations were carried out for the flow around a hydrodynamically smooth, fixed circular cylinder at two Reynolds numbers, one at a subcritical Reynolds number (Re = 1.4 × 105) and the other at a supercritical Reynolds number (Re = 1.0 × 106). The computations were made using a parallelized finite-volume Navier-Stokes solver based on a multidimensional linear reconstruction scheme that allows use of unstructured meshes. Central differencing was used for discretization of both convection and diffusion terms. Time-advancement scheme, based on an implicit, non-iterative fractional-step method, was adopted in conjunction with a three-level, backward second-order temporal discretization. Subgrid-scale turbulent viscosity was modeled by a dynamic Smagorinsky model adapted to arbitrary unstructured meshes with the aid of a test-filter applicable to arbitrary unstructured meshes. The present LES results closely reproduced the flow features observed in experiments at both Reynolds numbers. The time-averaged mean drag coefficient, root-mean-square force coefficients and the frequency content of fluctuating forces (vortex-shedding frequency) are predicted with a commendable accuracy.

1997 ◽  
Vol 119 (4) ◽  
pp. 219-225 ◽  
Author(s):  
X. Lu ◽  
C. Dalton ◽  
J. Zhang

A steady approach flow around a circular cylinder is investigated by using a large eddy simulation (LES) with the Smagorinsky subgrid-scale model. A second-order accurate in time fractional-step method and a combined finite-difference/spectral approximation are employed to solve the filtered three-dimensional incompressible Navier-Stokes equations. To demonstrate the viability and accuracy of the method, we present results at Reynolds numbers of 100, 3 × 103, 2 × 104, and 4.42 × 104. At Re = 100, the physical flow is two-dimensional and the calculation is done without use of the LES method. For the higher values of Re, the flow in the wake is three-dimensional and turbulent and the LES method is necessary to describe the flow accurately. Calculated values of lift and drag coefficients and Strouhal number are in good agreement with the experimentally determined values at all of the Reynolds numbers for which calculation was done.


1997 ◽  
Vol 119 (3) ◽  
pp. 519-525 ◽  
Author(s):  
Xiyun Lu ◽  
Charles Dalton ◽  
Jianfeng Zhang

Three-dimensional sinusoidally oscillating flows around a circular cylinder are investigated by using a viscous flow method (VFM) and a large eddy simulation (LES). A second-order accurate in time fractional step method and a combined finite-difference/spectral approximation are employed to solve the filtered incompressible Navier-Stokes equations. To demonstrate the viability and accuracy of the method, we calculate two cases of steady approach, flows at Reynolds numbers Re = 100 using VFM and Re = 104 using LES. For sinusoidally oscillating flows at β = 1035, the flow is 2D for KC< 0.5, 3D for 0.5 < KC < 2, and turbulent for KC > 2. For KC = 0.5, 0.8 and 1, the flow is calculated using VFM. For KC = 2, 3, 4, 5, 8 and 10, we have simulated the flow using LES with the Smagorinsky subgrid scale model. The drag and inertia coefficients are calculated from the in-line force acting on the cylinder and are in very good agreement with experimental data.


Author(s):  
Michael Leschziner ◽  
Ning Li ◽  
Fabrizio Tessicini

This paper provides a discussion of several aspects of the construction of approaches that combine statistical (Reynolds-averaged Navier–Stokes, RANS) models with large eddy simulation (LES), with the objective of making LES an economically viable method for predicting complex, high Reynolds number turbulent flows. The first part provides a review of alternative approaches, highlighting their rationale and major elements. Next, two particular methods are introduced in greater detail: one based on coupling near-wall RANS models to the outer LES domain on a single contiguous mesh, and the other involving the application of the RANS and LES procedures on separate zones, the former confined to a thin near-wall layer. Examples for their performance are included for channel flow and, in the case of the zonal strategy, for three separated flows. Finally, a discussion of prospects is given, as viewed from the writer's perspective.


Author(s):  
Francisco Jose´ de Souza ◽  
Aristeu Silveira Neto

Subgrid-scale modeling, which characterizes Large Eddy Simulation (LES), has been used to predict the behavior of a water-fed hydrocyclone operating without an air core. The governing equations were solved by a fractional step method on a staggered grid. The Smagorinsky subgrid-scale model was employed to account for turbulent effects. Numerical results actually capture the main features of the flow pattern and agree reasonably well with experiments, suggesting that LES represents an interesting alternative to classical turbulence models when applied to the numerical solution of fluid flows within hydrocyclones.


2012 ◽  
Vol 232 ◽  
pp. 471-476 ◽  
Author(s):  
Rui Zhao ◽  
Chao Yan

The flow past a circular cylinder at a subcritical Reynolds number 3900 was simulated by the method of detached-eddy simulation (DES). The objective of this present work is not to investigate the physical phenomena of the flow but to study modeling as well as numerical aspects which influence the quality of DES solutions in detail. Firstly, four typical spanwise lengths are chosen and the results are systematically compared. The trend of DES results along the span increment is different from previous large-eddy simulation (LES) investigation. A wider spanwise length does not necessary improve the results. Then, the influence of mesh resolution is studied and found that both too coarse and over refined grids will deteriorate the performance of DES. Finally, different orders of numerical schemes are applied in the inviscid fluxes and the viscous terms. The discrepancies among different schemes are found tiny. However, the instantaneous flow structures produced by 5th order WENO with 4th order central differencing scheme are more abundant than the others. That is, for the time-averaged quantities, the second-order accurate schemes are effective enough, whereas the higher-order accurate methods are needed to resolve the transient characteristics of the flow.


Author(s):  
Gorazd Medic ◽  
Om Sharma

Flow over three low-pressure turbine airfoils presented in [1] is analyzed for a range of Reynolds numbers (30,000 to 150,000) by means of large-eddy simulation. Baseline computational grid for these 2D linear cascade configurations consisted of 35 millions cells, and additional finer grids of 70 millions cells were used for grid sensitivity studies. For these low Reynolds number flows, this represents a quasi-DNS resolution which minimizes the role of the subgrid-scale model — however, WALE subgrid-scale model [7] was still employed. The configurations were analyzed for low free-stream turbulence intensity, as well as for 4% turbulence intensity at free-stream. Laminar separation exists on the suction side, and, depending on the Reynolds number, the flow at the outer edge of the separation either transitions, and the separation closes before the trailing edge, or not. Detailed comparisons to measurements are presented for computed surface pressure and total pressure losses over the range of Reynolds numbers for all three airfoils; these show that LES analyses are able to capture the main trends across all three geometries.


1990 ◽  
Vol 220 ◽  
pp. 459-484 ◽  
Author(s):  
H. M. Badr ◽  
M. Coutanceau ◽  
S. C. R. Dennis ◽  
C. Ménard

The unsteady flow past a circular cylinder which starts translating and rotating impulsively from rest in a viscous fluid is investigated both theoretically and experimentally in the Reynolds number range 103 [les ] R [les ] 104 and for rotational to translational surface speed ratios between 0.5 and 3. The theoretical study is based on numerical solutions of the two-dimensional unsteady Navier–Stokes equations while the experimental investigation is based on visualization of the flow using very fine suspended particles. The object of the study is to examine the effect of increase of rotation on the flow structure. There is excellent agreement between the numerical and experimental results for all speed ratios considered, except in the case of the highest rotation rate. Here three-dimensional effects become more pronounced in the experiments and the laminar flow breaks down, while the calculated flow starts to approach a steady state. For lower rotation rates a periodic structure of vortex evolution and shedding develops in the calculations which is repeated exactly as time advances. Another feature of the calculations is the discrepancy in the lift and drag forces at high Reynolds numbers resulting from solving the boundary-layer limit of the equations of motion rather than the full Navier–Stokes equations. Typical results are given for selected values of the Reynolds number and rotation rate.


2009 ◽  
Vol 620 ◽  
pp. 383-411 ◽  
Author(s):  
JUNGWOO KIM ◽  
HAECHEON CHOI

In the present study, the effects of the jet inflow conditions such as the initial momentum thickness (θ) and background disturbances on the downstream evolution of a circular jet are investigated using large eddy simulation (LES). We consider four different initial momentum thicknesses,D/θ = 50, 80, 120 and 180, and three different Reynolds numbers,ReD=UJD/ν = 3600, 104and 105, whereUJis the jet inflow velocity andDis the jet diameter. The present study shows that the jet characteristics significantly depend on both the initial momentum thickness and the Reynolds number. For all the Reynolds numbers considered in this study, vortex rings are generated at an earlier position with decreasing initial momentum thickness. In case of a relatively low Reynolds number ofReD= 3600, at smaller initial momentum thickness, early growth of the shear layer due to the early generation of vortex ring leads to the occurrence of large-scale coherent structures in earlier downstream locations, which results in larger mixing enhancement and more rapid increase in turbulence intensity. However, at a high Reynolds number such asReD= 105, with decreasing initial momentum thickness, rapid growth of the shear layer leads to the saturation of the shear layer and the generation of fine-scale turbulence structures, which reduces mixing and turbulence intensity. With increasingReθ(=UJθ/ν), the characteristic frequency corresponding to the shear layer mode (Stθ=fθ/UJ) gradually increases and reaches near 0.017 predicted from the inviscid instability theory. On the other hand, the frequency corresponding to the jet-preferred mode (StD=f D/UJ) varies depending onReDandD/θ. From a mode analysis, we show that, in view of the energy of the axial velocity fluctuations integrated over the radial direction, the double-helix mode (mode 2) becomes dominant past the potential core, but the axisymmetric mode (mode 0) is dominant near the jet exit. In view of the local energy, the disturbances grow along the shear layer near the jet exit: for thick shear layer, mode 0 grows much faster than other modes, but modes 0–3 grow almost simultaneously for thin shear layer. However, past the potential core, the dominant mode changes from mode 0 near the centreline to mode 1 and then to mode 2 with increasing radial direction regardless of the initial shear layer thickness.


Sign in / Sign up

Export Citation Format

Share Document