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 GR: Turbulent Boundary Layers: Structure & Pressure Gradient |
Hide Abstracts |
Chair: Karen Flack, United States Naval Academy Room: Hilton Chicago Stevens 3 |
Monday, November 21, 2005 10:34AM - 10:47AM |
GR.00001: Structure of Canonical Turbulent Wall-Bounded Flows Matthias H. Buschmann, Mohamed Gad-el-Hak To properly describe wall-bounded turbulent flows, a general idea on the structure of this type of flow is needed. In this talk, we focus on two-dimensional channel flows and zero-pressure-gradient boundary layers. Expanding on the idea advanced by Klewicki et al.\ ({\em J.\ Fluid Mech.} {\bf 222}, pp.\ 303--327, 2004), we analyze the continuity, mean momentum and transport equation of turbulence kinetic energy using channel DNS data. The outcome of this analysis is that the classical two-layer approach is physically most convincing and, for practical purposes (e.g., derivation of mean-velocity profile), most efficient approach. Several scaling schemes have been suggested based on the two-layer idea. In this presentation, we apply the Zagarola--Smits scaling---originally proposed based on empirical ground for the outer representation of pipe flows---to different types of canonical wall-bounded flows. A new suggestion to apply this scaling for the inner representation of the mean-velocity profile is made and successfully applied. [Preview Abstract] |
Monday, November 21, 2005 10:47AM - 11:00AM |
GR.00002: A Comparison of Turbulence within and above a Mature Corn Canopy to that of a Model Canopy. W. Zhu, R. Van Hout, J. Katz, L. Luznik, H.-S. Kang, C. Meneveau The turbulent flow structure within and above a mature corn canopy and a wind tunnel model canopy were measured using PIV. The Taylor scale Reynolds numbers ranged from 2000 to 3000 in the field, and from 500 to 800 in the tunnel. Measurements were performed at elevations ranging between 0.7 to 2.2 canopy heights. The model mean velocity profile was matched with the field data using variable screens. However, there were significant differences in Reynolds stress at the canopy height. In the field data we combine spatial and time series to obtain spectra ranging six decades of wave numbers, with an inertial range spanning three orders of magnitude. Using dissipation estimates for scaling, the laboratory and field data collapse onto a single curve that agrees with sonic anemometer data in the field, and hot-wire data in the laboratory. Quadrant-Hole analysis was used to study the flow during ejection and sweep events. Conditionally sampled momentum flux, transverse vorticity magnitude and dissipation rate reveal that sweep events dominate below and around canopy height, while ejections take over well above the canopy. Close to the canopy, especially for field data, quadrant one events contribute significantly to the mean vorticity magnitude and dissipation rate. [Preview Abstract] |
Monday, November 21, 2005 11:00AM - 11:13AM |
GR.00003: Scale by scale budgets for PIV and DNS data of wall-bounded turbulence N. Marati, E. Longmire, N. Saikrishnan, I. Marusic, C.M. Casciola, R. Piva The dynamics of the wall region at different scales recently has been analyzed for DNS of a channel at small Reynolds number through a generalized form of the K\'{a}rm\'{a}n-Howarth equation. The present work analyzes the scale-by-scale dynamics of wall-bounded flows at higher Re using both experimental and numerical data. The former data are taken from a dual-plane PIV experiment (that allows evaluation of the entire velocity gradient tensor) performed in a zero-pressure-gradient turbulent boundary layer at friction Reynolds number $Re_\tau = 1160$. The numerical data are taken from DNS of a turbulent channel flow by Zandonade and Moser, realized at a comparable $Re_\tau = 950$. Wall-normal locations in the log and wake regions were analyzed. In the log region, both data sets show the existence of a large-scale production range, followed by a nearly classical transfer range, closed by diffusion at the local dissipative scales. In the wake region the turbulent fluctuations are sustained by the spatial flux of scale-energy, while the local production plays a minor role. [Preview Abstract] |
Monday, November 21, 2005 11:13AM - 11:26AM |
GR.00004: An investigation of inner and outer-scaled structures in turbulent boundary layers. Ivan Marusic, Nicholas Hutchins The interaction between the large outer-scaled events in the logarithmic region and the near-wall buffer region structure will be investigated. The latter consists of the now quite well defined near-wall cycle of streaks and vortices (with a dominant spanwise spacing of 100 wall units), whilst the former has been recently noted to consist of very long meandering regions of positive and negative $u$ fluctuations (frequently exceeding $20$ boundary layer thicknesses in length, and with it's own associated vortical structure). Interestingly, these large outer-scaled structures maintain a footprint on the wall, seeming to modulate the near-wall cycle. This provides a mechanism for the percolation of very low wavenumber energy from the log region into the near-wall $u$ fluctuations, explaining the Reynolds number dependence of both the streamwise energy spectra $\Phi_{uu}$ and the peak of the RMS fluctuations in the near-wall region. Recent hotwire rake measurements will be presented along with results from direct numerical simulations. [Preview Abstract] |
Monday, November 21, 2005 11:26AM - 11:39AM |
GR.00005: Evaluation of Turbulence Models Through Predictions of a Simple 3D Boundary Layer. A. Jammalamadaka, K. Chauhan, H. Nagib Although a number of popular turbulence models are now commonly used to predict complex 3D flows, in particular for industrial applications, very limited full evaluation of their performance has been carried out using thoroughly documented experiments. One such experiment is that of Bruns, Fernholz and Monkewitz (JFM, vol. 393; 1999) in a boundary layer on the wall of an S-shaped duct, where the wall shear stress was measured accurately and independently in the original work and more recently with oil-film interferometry by Reudi et al. (Exp Fluids vol. 35; 2003). Results from various models including $k-\varepsilon $, Spalart-Alamaras, $k-\omega $, Menter's SST, and RSM are compared with the experimental results to extract better understanding of strengths and limitations of the various models. In addition to the various pressure distributions along the S-duct and the shear stress development on the test surface, the various normal stresses are compared for all the models with some surprising results in reference to the difficulty in predicting even such a simple 3D turbulent flow. Comparisons of other Reynolds stresses with models that predict them directly also reveal interesting results. In general the predictions of models are more in agreement with each other than with the experiment, suggesting that they suffer from common shortcomings. Also, the deviations of the predictions from the experiment grow to significant levels just beyond the development of the cross-over transverse velocity profile. [Preview Abstract] |
Monday, November 21, 2005 11:39AM - 11:52AM |
GR.00006: Intermediate shear flow states and the control of turbulence Fabian Waleffe, Jue Wang Shear flows are observed in two states: laminar and turbulent. But what lies in-between, on the boundary between the basins of attraction of those two states? We show that there exist remarkable quasi-2D traveling wave states that consist of large scale streaks with O(1) spanwise modulation in the streamwise velocity sustained by large scale O(1/R) streamwise rolls and a single marginally stable streak instability eigenmode. The latter has a 2D critical layer structure. These states are directly related to the transition threshold, which is defined as the amplitude of the smallest perturbation that triggers transition to turbulence. The drag induced by these quasi-2D states is typically only 10 to 40 percent higher than the {\it laminar} drag, asymptotically as R $\to \infty$, in contrast, the ratio of the turbulent to laminar drag diverges with R. The quasi-2D states have clear and few modes of instability, and clear, large scale structure. Hence, we propose that these states offer a new and promising avenue for the control of shear turbulence in addition to their importance for theoretical understanding of transition and, together with their small scale `upper branch' cousins, of the nature of shear turbulence. [Preview Abstract] |
Monday, November 21, 2005 11:52AM - 12:05PM |
GR.00007: Asymptotic Structure of Turbulent Boundary Layers: Multi-Valued Solutions and Boundary Layer Separation Bernhard Scheichl, Alfred Kluwick A comprehensive theory of incompressible turbulent boundary layers (TBL) is established by adopting a minimum of assumptions regarding the flow physics and properly investigating the equations of motion in the high- Reynolds-number limit. The logarithmic law of the wall is revealed and shown to be associated with an asymptotically small rotational streamwise velocity defect on top of the viscous wall layer. Consequently, the classical scaling of two-tiered TBL provides the simplest feasible flow structure. It is, however, possible to extend this concept and to formulate a three-tiered TBL having a slightly larger, i.e.\ a `moderately' large, velocity defect. This allows for, amongst others, the prediction of the in the past intensely debated phenomenon of non-unique equilibrium flows for a given pressure gradient. Most important, the observation that all commonly employed closures contain small numbers which may serve to measure the slenderness of the shear layer finally leads to a fully self-consistent asymptotic description of TBL which exhibit a velocity defect of \mbox{$O(1)$} and, in turn, can even undergo marginal separation. [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