Bulletin of the American Physical Society
2005 58th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 20–22, 2005; Chicago, IL
Session EP: Turbulence Simulations: DNS II |
Hide Abstracts |
Chair: P.K Yeung, Georgia Institute of Technology Room: Hilton Chicago Stevens 1 |
Sunday, November 20, 2005 4:10PM - 4:23PM |
EP.00001: Master-modes of the 3D turbulent channel flow Maksym Bondarenko, Sergei Chernyshenko Using Chebychev-Fourier representation of Direct Numerical Simulation solution we determine the so-called master modes, that is those modes which contain all essential information about the flow. The method used by E. Olson and E.S. Titi for 2D case is applied for 3D turbulent channel flow (i.e. Determining modes for continuous data assimilation in 2D turbulence, Journal of Statistical Physics, 113 (2003), 799-840). Initial simulation performed with 32786 Chebychev-Fourier modes using a spatial domain with streamwise and spanwise periods of 1.6 $\pi$ revealed that the number of master-modes for $Re_\tau$=85 is $N \leq 650$. Number of master-modes is not the same as, but may be related to, the fractal dimension of the attractor. For the comparison, L. Keefe, J. Kim and P. Moin estimated the fractal dimension as $D_\lambda$=780 for $Re_\tau$=80. (i.e. The dimension of attractors underlying periodic turbulent Poiseuille flow, J. Fluid Mech (1992), vol. 242, pp.1-29). Results for higher $Re_\tau$ will be obtained, analysed and reported at the conference. In particular we are interested in what organised structures will appear in the master modes. [Preview Abstract] |
Sunday, November 20, 2005 4:23PM - 4:36PM |
EP.00002: Karhunen-Lo\`{e}ve Eigenfunction Decomposition of Drag Reduced Turbulent Pipe Flow Results from a Spectral Element Direct Numerical Simulation Kenneth Ball, Andrew Duggleby Results of a Karhunen-Lo\`{e}ve Eigenfuncion Decomposition of the Direct Numerical Simulation flow field for both a fully turbulent pipe flow and a drag reduced pipe flow through spanwise wall oscillation will be presented. The flow field is decomposed into the eigenfunctions and eigenvalues of the two-point spatial correlation tensor for each azimuthal and axial wavenumber, and the energy and structure of the eigenfunctions are compared between the two flows to identify those structures affected by the wall oscillations and the mechanism responsible for drag reduction. The dynamical contribution of each eigenfunction mode, determined by orthogonal projection with the turbulent flow field, will also be examined. The flow field data is generated using NEK5000, a spectral element Navier-Stokes solver, where the polar-cylindrical coordinate singularity is avoided by solving the flow in Cartesian coordinates with a stadium-like element cross-section. Near the center of the pipe, a Cartesian configuration is used, while near the wall, the elements are mapped to a polar configuration. Each element uses 10th order Legendre Lagrangian interpolants in each direction, with a local Jacobi/Conjugate Gradient solver and a global Schwarz Multigrid solver. The flow field is generated for $Re_{\tau}=150$ using 2560 elements and a length of 20 R. [Preview Abstract] |
Sunday, November 20, 2005 4:36PM - 4:49PM |
EP.00003: Influence of solution rheology on the extent of polymer induced drag reduction in turbulent channel flow: A direct numerical simulation (DNS) study Chang-Feng Li, Radhakrishna Sureshkumar, Bamin Khomami Hi-fidelity DNS channel flow simulations of polymer induced drag reduction up to the maximum drag reduction (MDR) limit have been performed using a fully spectral method in conjunction with a number of kinetically theory based elastic dumbbell models for description of the polymer chain dynamics. The simulation results in turn have been used to develop a scaling that describes the interplay between fluid rheology (i.e., maximum chain extension and fluid relaxation time) and the extent of drag reduction as a function of Reynolds number. In addition, turbulence statistics are analyzed and correlations between the polymer body force, velocity fluctuations and vortical structures have been developed with particular emphasis on the high drag reduction (HDR) and the MDR regime. Based on these observations a mechanism for polymer induced drag reduction as well as an eddy viscosity model is proposed. [Preview Abstract] |
Sunday, November 20, 2005 4:49PM - 5:02PM |
EP.00004: Extraction of coherent vortices from high resolution DNS of homogeneous isotropic turbulence Marie Farge, Kai Schneider, Katsunori Yoshimatsu, Naoya Okamoto, Yukio Kaneda We have proposed a wavelet-based algorithm to extract coherent vortices out of turbulent flows. Since there is not yet a well-accepted definition of coherent structures for 3D flows, we suppose that they are what remains after denoising. Our {\it prior} is not on the structures themselves but on the noise, that we assume, as the simplest hypothesis, to be Gaussian and white. We apply this algorithm to several 3D homogeneous and isotropic turbulent flows forced at large scale and computed by DNS for different Taylor micro-scale Reynolds numbers, ranging from $R_\lambda= 167$ with resolution $N= 256^3$ to $R_\lambda=732$ with resolution $N=2048^3$. We found that the compression rate increases with $R_\lambda$, {\it i.e.}, the number of coefficients necessary to represent the coherent vortices drops from $3.6 \% N$ for $R_\lambda= 167$ to $2.6 \% N$ for $R_\lambda= 732$. The coherent vortices thus extracted contribute to about $99 \% $ of the total energy and about $80 \% $ of the total enstrophy. The corresponding coherent energy spectrum has the same $k^{-5/3}$ power-law behavior as the total energy, which corresponds to long-range correlation. In contrast, the incoherent energy scales in $k^{+2}$, which corresponds to decorrelation. We conjecture that discarding the incoherent flow is sufficient to model turbulent dissipation, as done in CVS (Coherent Vortex Simulation, see http://wavelets.ens.fr). [Preview Abstract] |
Sunday, November 20, 2005 5:02PM - 5:15PM |
EP.00005: Multiscale space--time adaptive simulation of 2D incompressible turbulence Jahrul Alam, Nicholas Kevlahan, Oleg Vasilyev A space--time adaptive wavelet collocation method is developed to efficiently simulate two-dimensional incompressible turbulence. This new DNS technique takes advantage of the spatial and temporal intermittency of turbulence to approximate the solution in the space--time domain using an adaptive collocation wavelet method. Both spatial and temporal resolution are adapted locally to solve the vorticity equation to the desired tolerance. Note that the global time integration error is controlled: this is not possible using conventional time marching methods. We will present results for the merging of identical vortices at Re = 1000, and for decaying two-dimensional turbulence. We find that the total number of active space--time degrees of freedom is significantly smaller than in a conventional dynamically adaptive time marching method. We also expect to present an estimate of the number of space--time degrees of freedom for decaying 2D turbulence as a function of Reynolds [Preview Abstract] |
Sunday, November 20, 2005 5:15PM - 5:28PM |
EP.00006: Direct Numerical Simulation of Turbulent Flow in a Wavy Channel Using an Efficient, Novel, Spectral Method Luo Wang, Kostas Housiadas, Peter Wapperom, Antony Beris A spectrally preconditioned biconjugate gradient algorithm has been developed to perform efficiently Direct Numerical Simulations (DNS) of Newtonian turbulent flow in a wavy channel. A transformation involving the shear direction only is applied to map the wavy geometry into a rectangular one so that a spectral approximation can be applied. DNS of Newtonian turbulent flow over a single sinusoidal wavy wall has been investigated at with the amplitude, wavelength of the undulation equal to 0.1, 2, respectively, based on the halfwidth of the channel. A fully implicit second order time integration scheme has been used with dealiasing along the periodic directions for all the non-linear terms using an influence matrix method to ensure the satisfaction of the divergence-free velocity condition (Housiadas and Beris 2004). The initial guess has been developed from the turbulent velocity profile obtained for a straight channel through the intermediate use of a pseudoconformal orthogonal coordinate system. The numerical results are compared against the experimental results reported in the literature (Hudson 1993). The influence of the wave amplitude on the structure of the turbulence is going to be discussed. [Preview Abstract] |
Sunday, November 20, 2005 5:28PM - 5:41PM |
EP.00007: Simulations of airfoil static and dynamic stall Santhanam Nagarajan, Sanjiva Lele Simulations of separated flow over stalled airfoils are conducted with an aim to understand post stall flow including separation and transition. A high-order accurate numerical methodology in curvilinear coordinates, along with overlapped zonal meshes is used to solve the compressible flow equations. The simulations resolve the boundary layer and are therefore a DNS in that region, while away from the airfoil, they reduce to LES. For a NACA 0012 airfoil at a high angle of attack ($15^{\circ}$) and low Reynolds number ($Re=135,000$), boundary layer separation is laminar, while breakdown to turbulence occurs through Kelvin-Helmholtz instabilities in the separated shear layer on the suction side of the airfoil. Boundary layer separation close to the leading edge leads to a significant region of recirculation where most of the turbulent fluctuations are concentrated. Turbulence escapes into the wake when larger vortices detach from the airfoil and convect downstream. The lift coefficient fluctuates in a chaotic manner, typical of stalled airfoils. A simulation of a pitching airfoil is being conducted to throw light on the phenomenon of dynamic stall and the ability of LES to predict large scale unsteady separation. Higher Reynolds number simulations, while not amenable to true LES, will be conducted using wall models. [Preview Abstract] |
Sunday, November 20, 2005 5:41PM - 5:54PM |
EP.00008: Aero-Optical Distortions by a Turbulent Wake Ali Mani, Meng Wang, Parviz Moin The aero-optical distortions caused by the turbulent wake behind a circular cylinder at $Re_D=3900$ and $M=0.4$ are investigated numerically. Large-eddy simulation is employed to compute the spatial and temporal variations of the index-of-refraction field, and a combination of ray tracing and Fourier optics is used to track the optical propagation and its far-field intensity patterns. Instantaneous and statistical descriptions of the optical aberrations are obtained for different flow resolution, optical wavelengths, and distances of propagation. An analytical description based on statistical solutions of the paraxial wave equation is provided to support the computed statistical behavior of beam propagation. Our results confirm that the effective range of an optical beam can be severely compromised by turbulence. In the parameter range considered, small scales of the flow are found to be optically active and must therefore be computationally resolved or modeled. It is found that the root-mean-square of the gradient of a distorted wave front plays a key role in causing beam spread as it propagates to the far field. [Preview Abstract] |
Sunday, November 20, 2005 5:54PM - 6:07PM |
EP.00009: Direct Numerical Simulation of the Thermal Effects of Plasmas on Turbulent Flows. Shankar Ghosh, Krishnan Mahesh The thermal effects of plasmas on isotropic turbulence are studied using direct numerical simulations. The turbulence is assumed to be spatially homogeneous and isotropic prior to generation of the plasma. Two idealizations of the plasma are considered - spherical and conical. The spherical idealization represents a point plasma. The conical idealization approximates the tear-drop shape of the plasma region that is observed experimentally. The plasma generates a blast wave which produces a toroidal region of vorticity for the tear-drop idealization. The variation of the magnitude of vorticity with temperature ratio and size of the plasma region is examined. The shock wave gets distorted as it interacts with the background turbulence. The turbulence is seen to be suppressed in the region occupied by the plasma and slightly amplifies across the blast wave. Details will be discussed. [Preview Abstract] |
Sunday, November 20, 2005 6:07PM - 6:20PM |
EP.00010: Large scale structures and energy transfer in hydrodynamic turbulence Pablo Mininni, Alexandros Alexakis, Annick Pouquet With the help of direct numerical simulations, we investigate the transfer of energy and triadic interactions in fully developed forced three-dimensional hydrodynamic turbulence. The assumption of locality of transfer among the different scales is one of the building blocks of Kolmogorov (1941) theory of turbulence. We use simulations on a grid of $1024^3$ points of a flow forced with a Taylor-Green vortex. Reynolds numbers of $R=790$ (based on the Taylor lengthscale) are reached. In the steady state, the flow displays a well defined large scale pattern superimposed with turbulent fluctuations at small scales. We find that nonlinear triadic interactions are dominated by their non-local components, involving widely separated scales, even though the nonlinear transfer itself is local and the scaling for the energy spectrum is close to the classical Kolmogorov law. These non-local interactions with large scales represent 20\% of the total energy flux. The link between these findings and the intermittency of the small scales, and their consequences for modeling of turbulent flows are also briefly discussed. [Preview Abstract] |
Sunday, November 20, 2005 6:20PM - 6:33PM |
EP.00011: Linear forcing in numerical simulations of isotropic turbulence Carlos Rosales, Charles Meneveau Simulations of forced isotropic turbulence are most often formulated in Fourier space, where forcing is applied to low-wavenumber modes. That forcing is difficult to implement for applications in physical space. A linear forcing recently proposed by Lundgren, where a force proportional to velocity is applied, is an attractive alternative but not much is known about its properties. Using numerical experimentation, various properties of the linear forcing are explored. It is shown that when implemented in physical space linear forcing gives the same results as in spectral implementations and that the linearly-forced system converges to a stationary state that depends on domain size and Reynolds number, but not on the spectral shape of the initial condition. It is also shown that the extent of Kolmogorov -5/3 range is similar to that achieved using the standard band-limited forcing scheme but the integral length scale is smaller, reducing the effective scaling range for a given resolution. It is concluded that linear forcing is a useful alternative that does not require transformation to Fourier space and is easily integrated into physical-space numerical codes. [Preview Abstract] |
Sunday, November 20, 2005 6:33PM - 6:46PM |
EP.00012: Numerical simulations of bypass transition in flat-plate boundary layers Victor Ovchinnikov, Ugo Piomelli, Meelan M. Choudhari Numerical simulations of bypass transition in the Blasius boundary layer due to free-stream turbulence (FST) typically exclude the flat-plate leading edge. The inflow is usually placed in the Blasius boundary-layer region, and several assumptions are made about the state of the disturbances and the boundary layer at the arbitrary inflow location. We have performed Direct Numerical Simulations (DNS) of bypass transition due to high-amplitude FST in which the leadinge edge of the plate is included. The simulations start well upstream of the flat plate and extend into the fully turbulent region. In the three cases presented, we discuss the effects of varying FST intensity and length scale on the transition process, and examine the evolution of boundary-layer disturbances and the generation of turbulent spots. We present various mean-flow statistics, spectra, correlations, and flow visualizations, and draw comparisons with previous studies with lower FST amplitudes. Research sponsored by the NASA Langley Research Center. [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