0
Research Papers: Techniques and Procedures

An Investigation of the Lattice Boltzmann Method for Large Eddy Simulation of Complex Turbulent Separated Flow

[+] Author and Article Information
Kannan N. Premnath

Visiting Fellow
Mem. ASME
CUNY Energy Institute,
City College of New York,
City University of New York,
New York, NY 10031
e-mail: kannan.premnath@ucdenver.edu

Martin J. Pattison

HyPerComp Inc.,
2629 Townsgate Road,
Suite 105,
West Lake Village, CA 91361

Sanjoy Banerjee

CUNY Distinguished Professor
Mem. ASME
Director
CUNY Energy Institute,
Department of Chemical Engineering,
City College of New York,
City University of New York,
New York, NY 10031

1Corresponding author.

2Department of Mechanical Engineering, University of Colorado Denver, 1200 Larimer Street, Denver, CO 80217.

Manuscript received March 24, 2012; final manuscript received January 9, 2013; published online April 3, 2013. Assoc. Editor: Ali Beskok.

J. Fluids Eng 135(5), 051401 (Apr 03, 2013) (12 pages) Paper No: FE-12-1147; doi: 10.1115/1.4023655 History: Received March 24, 2012; Revised January 09, 2013

Lattice Boltzmann method (LBM) is a relatively recent computational technique for fluid dynamics that derives its basis from a mesoscopic physics involving particle motion. While the approach has been studied for different types of fluid flow problems, its application to eddy-capturing simulations of building block complex turbulent flows of engineering interest has not yet received sufficient attention. In particular, there is a need to investigate its ability to compute turbulent flow involving separation and reattachment. Thus, in this work, large eddy simulation (LES) of turbulent flow over a backward facing step, a canonical benchmark problem which is characterized by complex flow features, is performed using the LBM. Multiple relaxation time formulation of the LBM is considered to maintain enhanced numerical stability in a locally refined, conservative multiblock gridding strategy, which allows efficient implementation. Dynamic procedure is used to adapt the proportionality constant in the Smagorinsky eddy viscosity subgrid scale model with the local features of the flow. With a suitable reconstruction procedure to represent inflow turbulence, computation is carried out for a Reynolds number of 5100 based on the maximum inlet velocity and step height and an expansion ratio of 1.2. It is found that various turbulence statistics, among other flow features, in both the recirculation and reattachment regions are in good agreement with direct numerical simulation and experimental data.

FIGURES IN THIS ARTICLE
<>
Copyright © 2013 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Schematic illustration of the three-dimensional, nineteen velocity (D3Q19) model

Grahic Jump Location
Fig. 2

Local grid refinement when using a staggered arrangement

Grahic Jump Location
Fig. 3

Interface between coarse and fine grid when using a staggered arrangement

Grahic Jump Location
Fig. 4

Schematic representation of geometry. Step is at x/h = 0 and z = 0 corresponds to lower boundary. Note that dimensions of some regions are exaggerated to ensure clarity.

Grahic Jump Location
Fig. 5

Mean streamwise velocity profiles. From left to right, positions are x/h = 0.5, 1, 2.5, 4, 6, 10, 15, 19; successive profiles are offset by U0 on abscissa: (–) this study; (- -) DNS [6].

Grahic Jump Location
Fig. 6

Mean streamlines computed by the multiblock MRT-LBM using the dynamic SGS model. Note that vertical coordinate is exaggerated by a factor of two for clarity.

Grahic Jump Location
Fig. 7

Instantaneous values of the vorticity computed by the multiblock MRT-LBM using the dynamic SGS model for three different representative times

Grahic Jump Location
Fig. 8

Turbulence intensities and Reynolds stress downstream of the step at different stations: (–) multiblock MRT-LBM using the dynamic SGS model; (- -) DNS [6]; (o) experiment [7]

Grahic Jump Location
Fig. 9

Comparison of turbulence intensities and Reynolds stress close to the corner in the recirculation region at x/h = 2: (–) multiblock MRT-LBM using the dynamic SGS model; (o) DNS [5]; (+) LES [5]; (×) coarse DNS [5]

Grahic Jump Location
Fig. 10

Comparison of the variation of the step-wall pressure coefficient: (–) multiblock MRT-LBM using the dynamic SGS model; (- -) DNS [6]; (o) experiment [7]

Grahic Jump Location
Fig. 11

Mean velocity profile in a fully developed turbulent channel flow at a shear Reynolds number Re*=183.7 plotted as a function of the wall normal distance normalized by viscous scales in semilog coordinates

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In