Large eddy simulation of a circular jet: effect of inflow conditions on the near field

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.

2014 ◽  
Vol 136 (6) ◽  
Author(s):  
Hieu T. Pham ◽  
Sutanu Sarkar

The performance of the large eddy simulation (LES) approach in predicting the evolution of a shear layer in the presence of stratification is evaluated. The LES uses a dynamic procedure to compute subgrid model coefficients based on filtered velocity and density fields. Two simulations at different Reynolds numbers are simulated on the same computational grid. The fine LES simulated at a low Reynolds number produces excellent agreement with direct numerical simulations (DNS): the linear evolution of momentum thickness and bulk Richardson number followed by an asymptotic approach to constant values is correctly represented and the evolution of the integrated turbulent kinetic energy budget is well captured. The model coefficients computed from the velocity and the density fields are similar and have a value in range of 0.01-0.02. The coarse LES simulated at a higher Reynolds number Re = 50,000 shows acceptable results in terms of the bulk characteristics of the shear layer, such as momentum thickness and bulk Richardson number. Analysis of the turbulent budgets shows that, while the subgrid stress is able to remove sufficient energy from the resolved velocity fields, the subgrid scalar flux and thereby the subgrid scalar dissipation are underestimated by the model.


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.


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.


2014 ◽  
Vol 136 (5) ◽  
Author(s):  
Trushar B. Gohil ◽  
Arun K. Saha ◽  
K. Muralidhar

A large eddy simulation (LES) of an incompressible spatially developing circular jet at a Reynolds number of 10,000 is performed. The shear-improved Smagorinsky model (Lévêque et al., 2007, “A Shear-Improved Smagorinsky Model for the Large-Eddy Simulation of Wall-Bounded Turbulent Flows,” J. Fluid Mech., 570, pp. 491–502) is used for the resolution of the subgrid stress tensor within the filtered three-dimensional unsteady Navier–Stokes equations. Higher-order spatial and temporal discretization schemes are used for capturing the details of the turbulent flow field. With the help of instantaneous and time-averaged flow data, the spatial transition from the laminar state to the turbulent is analyzed. Flow structures are visualized using isosurfaces of the Q-criterion. Instantaneous flow patterns show single tearing and multiple pairing processes. Tracing individual vortex rings over a longer time period, a detailed understanding of the vortex interaction is revealed. The observed trends and the length of the potential core are in conformity with the findings of earlier experiments. The time-averaged axial velocity profile shows that the jet attains self-similarity and the computed profile matches well with the experimental results of Hussein et al. (1994, “Velocity Measurements in a High-Reynolds-Number, Momentum-Conserving, Axisymmetric, Turbulent Jet,” J. Fluid Mech., 258, pp. 31–75). The centerline decay of the velocity and entrainment rate are in agreement with published experiments. The Reynolds stress components u'u'¯, v'v'¯, and u'v'¯ and the third-order velocity moment are in good agreement with thr experimental results, thus confirming the validity of the present simulation.


2014 ◽  
Vol 759 ◽  
pp. 579-611 ◽  
Author(s):  
Jae Hwan Jung ◽  
Hyun Sik Yoon

AbstractWe consider a twisted cylinder that was designed by rotating the elliptic cross-section along the spanwise direction, resulting in a passive control. The flow over the twisted cylinder is investigated at a subcritical Reynolds number (Re) of 3000 using large eddy simulation based on the finite volume method. For comparison, the flow past smooth and wavy cylinders is also calculated. The twisted cylinder achieves reductions of approximately 13 and 5 % in mean drag compared with smooth and wavy cylinders, respectively. In particular, the root mean square (r.m.s.) value of the lift fluctuation of the twisted cylinder shows a substantial decrease of approximately 96 % compared with the smooth cylinder. The shear layer of the twisted cylinder covering the recirculation region is more elongated than those of the smooth and wavy cylinders, and vortex shedding from the twisted cylinder is considerably suppressed. Consequently, the elongation of the shear layer from the body and the near disappearance of vortex shedding in the near wake with weak vortical strength contributes directly to the reduction of drag and lift oscillation. Various fundamental mechanisms that affect the flow phenomena, three-dimensional separation, pressure coefficient, vortex formation length and turbulent kinetic energy are examined systematically to demonstrate the effect of the twisted cylinder surface. In addition, for the twisted cylinder at $\mathit{Re}=3000$, the effect of the cross-sectional aspect ratio is investigated from 1.25 to 2.25 to find an optimal value that can reduce the drag and lift forces. Moreover, the effect of the Reynolds number on the aerodynamic characteristics is investigated in the range of $3\times 10^{3}\leqslant \mathit{Re}\leqslant 1\times 10^{4}$. We find that as Re increases, the mean drag and the r.m.s. lift coefficient of the twisted cylinder increase, and the vortex formation length decreases.


2011 ◽  
Vol 15 (2) ◽  
pp. 367-378 ◽  
Author(s):  
Zouhaier Mehrez ◽  
Mourad Bouterra ◽  
Cafsi El ◽  
Ali Belghith ◽  
Quere Le

The control of fluid mechanics and mass transfer in separated and reattaching flow over a backward-facing step by a local forcing, is studied using Large Eddy Simulation (LES). To control the flow, the local forcing is realized by a sinusoidal oscillating jet at the step edge. The Reynolds number is varied in the range 10000 ? Re ? 50000 and the Schmidt number is fixed at 1. The found results show that the flow structure is modified and the local mass transfer is enhanced by the applied forcing. The observed changes depend on the Reynolds number and vary with the frequency and amplitude of the local forcing. For the all Reynolds numbers, the largest recirculation zone size reduction is obtained at the optimum forcing frequency St = 0.25. At this frequency the local mass transfer enhancement attains the maximum.


2016 ◽  
Vol 138 (7) ◽  
Author(s):  
Ye Zhou ◽  
Ben Thornber

The implicit large-eddy simulation (ILES) has been utilized as an effective approach for calculating many complex flows at high Reynolds number flows. Richtmyer–Meshkov instability (RMI) induced flow can be viewed as a homogeneous decaying turbulence (HDT) after the passage of the shock. In this article, a critical evaluation of three methods for estimating the effective Reynolds number and the effective kinematic viscosity is undertaken utilizing high-resolution ILES data. Effective Reynolds numbers based on the vorticity and dissipation rate, or the integral and inner-viscous length scales, are found to be the most self-consistent when compared to the expected phenomenology and wind tunnel experiments.


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.


Sign in / Sign up

Export Citation Format

Share Document