Bulletin of the American Physical Society
2006 59th Annual Meeting of the APS Division of Fluid Dynamics
Sunday–Tuesday, November 19–21, 2006; Tampa Bay, Florida
Session KM: Turbulence Simulations IV |
Hide Abstracts |
Chair: P.K. Yeung, Georgia Institute of Technology Room: Tampa Marriott Waterside Hotel and Marina Meeting Room 10 |
Monday, November 20, 2006 5:15PM - 5:28PM |
KM.00001: Influence of Initial Conditions on Decaying Two-Dimensional Turbulence Ruben Trieling, Laurens van Bokhoven, Herman Clercx, GertJan van Heijst A numerical study of freely decaying two-dimensional turbulence is presented to show how the time evolution of characteristic flow quantities is influenced by the initial conditions. The numerical method adopted is a standard two-dimensional (2D) Fourier pseudo-spectral algorithm with Newtonian viscosity. Vortex statistics have been extracted using a vortex census method. Several characteristic initial vorticity distributions analogous to those employed in previous laboratory experiments are considered. All initial vorticity distributions have in common a dominant subset of vortices. Reliable statistics have been obtained for each characteristic distribution by ensemble averaging. The evolutions of the average wavenumber and the number density related to a dominant subset are found to collapse confirming the self-similarity of 2D turbulence, one of the starting points for the scaling theory proposed by Carnevale et al. [Phys. Rev. Lett. 66, 2735 (1991)]. The mutual ratios of the relevant scaling exponents as predicted by the scaling theory are not confirmed, however, and thus seem questionable within the considered parameter range. Furthermore, power-law exponents for both the number density and the average wavenumber are found to be affected by the initial number density. This suggests that for experiments in shallow fluid layers any agreement with a universal scaling exponent seems coincidental. [Preview Abstract] |
Monday, November 20, 2006 5:28PM - 5:41PM |
KM.00002: Parametric Dependence of Strain and Vorticity in Homogeneous Shear Flow on Reynolds Number and Shear Parameter Juan Isaza, Lance Collins The combined role of the Shear Parameter, $S^{*} = S k / \epsilon $, and the Reynolds Number in Homogeneous Shear Flow is studied using Direct Numerical Simulations (DNS). Particular attention is given to enstrophy, strain, their production and geometrical statistics. The parametric investigation involves DNS of $512^{3}$ with Shear Parameter values between 10 and 50 and Reynolds Number above 40. This study is motivated by some of the results reported by Kholmyansky \emph{et al}, [Phys Fluids 13, 2001]. The results at high values of the shear parameter are compared with the prediction of Rapid Distortion Theory. [Preview Abstract] |
Monday, November 20, 2006 5:41PM - 5:54PM |
KM.00003: Direct Numerical Simulations of Stably Stratified Turbulent Boundary Layers at High Re Junwoo Lim, Byung-Gu Kim, Changhoon Lee We performed direct numerical simulations (DNS) of stably stratified turbulent boundary layers, in the range of $Re_ {\tau} = 180 \sim 800$ and $Ri = 0 \sim 400$. To extend our investigation on the dynamics of near-wall turbulent structures under strong stratification to higher Reynolds numbers, we developed a highly scalable parallel program. Compared to the previous version, extensive improvements on the computation efficiency and the communication scalability were made so that tera-scale supercomputers could be fully exploited. Our simulations show the evidences of modification in the near-wall structures due to the stratification, as well as the typical suppression of turbulence far away from the wall as previously reported by other researchers. Interestingly, for some cases, DNS resulted in complete suppression of the near-wall disturbances while large-eddy simulations (LES) for the same cases did not. More detailed discussion on this issue and a few other computational issues associated with the increase of the Richardson number will be addressed in the presentation. [Preview Abstract] |
Monday, November 20, 2006 5:54PM - 6:07PM |
KM.00004: DNS of a passive scalar in a turbulent channel with local forcing at walls. Guillermo Araya, Stefano Leonardi, Luciano Castillo, Paolo Orlandi Direct Numerical Simulations (DNS) of the velocity and thermal fields in a fully developed turbulent channel, with normal periodic blowing/suction velocity disturbances at both walls, are presented. The governing equations have been discretized in an orthogonal coordinate system using a staggered central second-order finite-difference approximation. Results at low Reynolds number show a peak drag reduction of 60 percent and an average drag reduction of 46 percent with respect to the unperturbed channel when using a specific combination of amplitude/frequency in the local forcing system. Onward investigations consider the analysis of higher Reynolds numbers as well as influence of the local forcing on the heat transfer. [Preview Abstract] |
Monday, November 20, 2006 6:07PM - 6:20PM |
KM.00005: 3D Large-Scale DNS of Weakly-Compressible Homogeneous Isotropic Turbulence With Lagrangian Tracer Particles R. Fisher, D. Lamb, L. Kadanoff, F. Cattaneo, P. Constantin, T. Plewa When simulating turbulence with complex or embedded geometries, or which transitions from incompressible to weakly-compressible, it is desirable to have a robust numerical method which is equally capable of handling these regimes without significant loss of accuracy. The FLASH 2006 turbulence simulation is a driven, weakly-compressible, homogeneous, isotropic simulation which explores this concept in detail. It was performed at $1856^3$ resolution with 16.7 million Lagrangian tracer particles at a (1D) RMS velocity of 0.17. The simulation was performed by special invitation on the LLNL BG/L machine shortly before it was permanently placed inside their secure network earlier this year. Approximately one week of CPU time on 65,536 processors were used. We will present results including both Eulerian and Lagrangian properties of the simulation, and compare these to previous experiments and theories. We will also discuss a systematic error in the determination of the higher-order structure functions due to finite statistics and address this issue for our dataset. [Preview Abstract] |
Monday, November 20, 2006 6:20PM - 6:33PM |
KM.00006: High-resolution 2D and 3D numerical simulations of gravity currents Afshin Eghbalzadeh, Joongcheol Paik, Fotis Sotiropoulos Gravity currents are essentially horizontal motion of fluids of different density in a gravitational field and their spreading behaviors are characterized by the relative balance of buoyancy, inertial and viscous forces. High resolution numerical simulations of two- and three-dimensional (2D and 3D) gravity currents are carried out using the unsteady Reynolds-averaged Navier-Stokes (URANS) closed by the buoyancy-extended turbulence model. The governing equations are solved using second-order-accurate spatial and temporal discretization methods. Comparison with available experimental measurements confirms that high resolution URANS can capture reasonably well the rich dynamics of coherent vortical structures at the interface of the gravity current and the ambient flow, front speed of energy conserving head region and billow breakdown in wake region. Details of 2D and 3D numerical solutions of gravity currents with available experimental measurements and their flow physics will be presented at the conference. [Preview Abstract] |
Monday, November 20, 2006 6:33PM - 6:46PM |
KM.00007: Lagrangian conditional statistics, acceleration and local relative motion in direct numerical simulations of turbulence P.K. Yeung, S.B. Pope, E.A. Kurth, A.G. Lamorgese Lagrangian statistics of fluid particle velocity and acceleration conditioned on fluctuations of dissipation, enstrophy and pseudo-dissipation representing local relative motion in the flow are extracted from a direct numerical simulation (DNS) database of forced, stationary isotropic turbulence. The grid resolution is up to $2048^3$, and the Taylor-scale Reynolds number covers a range up to about 650. Conditional velocity autocorrelations are consistent with rapid changes for the velocity of fluid particles moving in regions of large velocity gradients. Autocorrelations conditioned on the enstrophy show distinctive features which are, through the use of coordinate axes parallel and perpendicular to the vorticity vector, traced to vortex-trapping effects studied by others in the literature. However, rapid changes in vorticity vector orientation make these effects weaker at high Reynolds numbers. Further results including conditional velocity-acceleration cross-correlations, which involve a degree of flow detail currently accessible only in DNS, are also used to help develop a new stochastic model that accounts for the effects of turbulence intermittency at the small scales. [Preview Abstract] |
Monday, November 20, 2006 6:46PM - 6:59PM |
KM.00008: Is the Kelvin Theorem Valid for High-Reynolds-Number Turbulence? M. Wan, Z. Xiao, S. Chen, G. Eyink The Kelvin-Helmholtz theorem on conservation of circulations is supposed to hold for ideal inviscid fluids and is believed to be play a crucial role in turbulent phenomena, such as production of dissipation by vortex line-stretching. However, this expectation does not take into account singularities in turbulent velocity fields at infinite Reynolds number. Recent theory [1,2] predicts a ``cascade of circulations'' due to the vortex-force induced by small-scale (subgrid) turbulence. We present evidence from a numerical simulation of the three-dimensional turbulent energy cascade [3] both for instantaneous violation of circulation conservation (non-zero time-derivative) and also for violation over finite intervals of time, by Lagrangian tracking of material loops. Although violated in individual realizations, we find that the circulations are still conserved in some average sense. For comparison, we show that Kelvin's theorem holds for individual realizations in the two-dimensional enstrophy cascade, in agreement with theory. We also verify quantitative expressions for the turbulent vortex-force, predicted by a multi-scale gradient expansion [2]. Our results provide the first clear evidence for breakdown of Kelvin's theorem (1869) in turbulent flow. Supported by NSF grant \# ASE-0428325 at the Johns Hopkins University. [1] G. L. Eyink, Comptes Rendus Physique, 7: 449-455 (2006). physics/0605014 [2] G. L. Eyink, Phys. Rev. E, submitted (2006). physics/0606159 [3] S. Chen et al., Phys. Rev. Lett., submitted (2006). physics/0605016 [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700