Bulletin of the American Physical Society
62nd Annual Meeting of the APS Division of Fluid Dynamics
Volume 54, Number 19
Sunday–Tuesday, November 22–24, 2009; Minneapolis, Minnesota
Session EA: Turbulent Boundary Layers: Structure |
Hide Abstracts |
Chair: Ronald Adrian, Arizona State University Room: 101A |
Sunday, November 22, 2009 4:15PM - 4:28PM |
EA.00001: A Method of Synthesizing 1-D POD Modes for Channel Flow Jon Baltzer, Ronald Adrian POD analysis is performed on DNS channel data for several Reynolds numbers between $Re_{\tau} = 180$ and $950$. The one-dimensional eigenmodes in the wall-normal direction are oscillatory and their phases can be extracted by various means. The phases contain the physics implied by the two-point spatial correlation, and the phases scaled using sequency correlate well with a single curve for all but the lowest mode numbers. An orthogonal basis to approximate the POD modes has been developed using this single phase function. Convergence using this basis is comparable to POD and superior to conventional orthogonal polynomials. The phase function is also compared to asymptotic results for large mode number (Moser, Phys. Fluids, \textbf{6}, 794-801, 1994). [Preview Abstract] |
Sunday, November 22, 2009 4:28PM - 4:41PM |
EA.00002: Energetic modes in turbulent pipe flow from resolvent analysis Ati Sharma, Beverley Mckeon We describe a method to investigate the mode shapes in turbulent pipe flow at a given wavenumber pair that are most responsive to harmonic forcing. Specifically, these modes correspond to the largest singular values in a Schmidt decomposition of the linear Navier-Stokes operator using the turbulent mean profile as the base flow. The ideas follow logically from the work of Sharma \& McKeon (2009), who considered a similar approach for laminar pipe flow. The dominant modes exhibit the inner and outer scaling behavior expected from experiment. A comparison of the mode shapes with the largest response with the results of a Karhunen-Loeve analysis on a pipeflow DNS by Duggleby \textit{et al} (2007), i.e. an \textit{a posteriori} evaluation of the most energetic mode shapes, shows reasonable agreement. An additional consequence of this analysis in terms of the scaling of the very large scale motions is discussed further in the presentation by McKeon \& Sharma. [Preview Abstract] |
Sunday, November 22, 2009 4:41PM - 4:54PM |
EA.00003: Conditional structure of large- and small-scales in high Reynolds number turbulent boundary layers B. Ganapathisubramani, N. Hutchins, J.P. Monty, I. Marusic A spanwise array of surface hot-film shear-stress sensors and a traversing hot-wire located directly above one of these sensors are used to identify the conditional structure of a high Reynolds number turbulent boundary layer. The shear-stress data is low-pass filtered and is used to detect large-scale low and high skin-friction events. The velocity fluctuations from the hot-wire are decomposed in to large-scale and small-scale components and used to compute conditional large-scale mean velocity and small-scale turbulence intensities conditioned on the presence of large-scale low and high skin-friction events (detected by the skin-friction sensors). The conditional mean based on the large-scales show a forward-leaning low- and high-speed structure for low and high skin-friction events. The conditional small-scale turbulence intensity is weak near the wall and intense farther away from the wall for a low skin-friction event and vice-versa for a high skin-friction event. The changes in the velocity gradient of the conditional large-scale structure is used to explain the trends in the small-scale activity. [Preview Abstract] |
Sunday, November 22, 2009 4:54PM - 5:07PM |
EA.00004: DNS study of very-long coherent structures in turbulent pipe flow Ronald Adrian, Xiaohua Wu, Jon Baltzer Fully developed incompressible turbulent pipe flow at Reynolds number $Re_D=24,580$ is simulated with second-order finite differences on a streamwise-periodic, 536 million point grid. The Karman number $R^+=648.8$, and the computational domain length is $30R$. The mean, second-order statistics and two-point correlations agree well with published experimental data. Pre-multiplied power spectra of the streamwise velocity peak at two wave numbers, one corresponding to very large scale motions (VLSM) $3\mbox{--}15R$ long, and the other corresponding to large scale motions (LSM) less than $3R$ long, consistent with earlier work (Kim and Adrian, Phys. Fluids \textbf{2}, 417--422, 1999, \textit{et seq.}). The low speed patterns of VLS motion convect at the bulk velocity, and they are associated with strong, radially inward ejections from layers closer to the wall. They are made up of smaller structures having spacings of $\sim0.2R=125^+$ or less, coinciding with the typical spacing between hairpins in a near wall packet. The VLS motions also appear to leave a clear low-speed pattern in the near wall regions. [Preview Abstract] |
Sunday, November 22, 2009 5:07PM - 5:20PM |
EA.00005: The Structure of Large- and Very Large-Scale Motions in Turbulent Pipe Flow Alexander Smits, Sean Bailey Multi-point velocity measurements in turbulent pipe flow have been performed at $Re_D=1.5 \times 10^5$. Using cross-spectral and Proper Orthogonal Decomposition analysis, information is elucidated on the structure of the large- and very-large scale motions in the outer layer of wall-bounded flows. The results indicate that the large-scale motions are composed of attached eddies in the logarithmic layer but are mostly composed of detached eddies in the outer layer with a wide range of azimuthal scales. The very-large-scale motions have large radial and azimuthal scales, are concentrated around a single azimuthal mode, and make a smaller angle with the wall compared to the large-scale motions. The results support a hypothesis that only the detached large-scale motions in the outer layer align to form the very-large-scale motions. [Preview Abstract] |
Sunday, November 22, 2009 5:20PM - 5:33PM |
EA.00006: Identification and analysis of boundary layer structures in Tomographic PIV data Cecilia Ortiz-Duenas, Neelakantan Saikrishnan, Ellen Longmire Tomographic particle image velocimetry data were acquired in a turbulent boundary layer flow generated in a wind tunnel facility. The friction Reynolds number \textit{Re}$_{\tau }$ was 1160. Measurement volumes of streamwise and spanwise dimension 0.7\textit{$\delta $} and wall normal thickness 0.07\textit{$\delta $}, which resolved the range $z^{+}$ = 70-150, were analyzed using custom software. Various criteria, including streamwise velocity, two- and three-dimensional swirl, individual vorticity components, and Reynolds shear stress, in combination with region growing and coincidence algorithms, were employed to identify and characterize coherent structures present in instantaneous fields. The results of the present study will be described and also compared with results from earlier experiments by Ganapathisubramani et al., 2003, 2006, that relied on planar PIV data only. [Preview Abstract] |
Sunday, November 22, 2009 5:33PM - 5:46PM |
EA.00007: Hairpin packet structure of a turbulent boundary layer in inclined wall-normal/spanwise planes Jae Hwa Lee, Hyung Jin Sung Turbulent coherent structures associated with hairpin packet motions have been scrutinized using the instantaneous flow fields obtained from the direct numerical simulation (DNS) of a turbulent boundary layer (TBL). The Reynolds number based on the momentum thickness was varied in the range \textit{Re}$_{\theta }$=890$\sim $2560. This study focused on the hairpin packet motions in inclined wall-normal/spanwise planes. The hairpin vortex signature associated with the hairpin leg components in the vertical inclined plane consists of a counter-rotating vortex pair, upward and downward motions and a stagnation point induced by the Q2 and Q4 events. These hairpin signatures were observed in the instantaneous flow field, in the two-point correlations and in the conditionally averaged flow fields, respectively. We considered three inclined planes (45$^{\circ}$, 90$^{\circ}$, and 135$^{\circ})$ to investigate the spatial characteristics of the hairpin packet motions in the log and wake regions. The statistical flow fields showed that significantly different flow patterns are induced by the intersections of the three inclined planes with the hairpin packet motions. [Preview Abstract] |
Sunday, November 22, 2009 5:46PM - 5:59PM |
EA.00008: On the Existence of Hairpin Vortices Peter Bernard Numerical simulation of the transitioning and turbulent boundary layer using a vortex filament scheme (AIAA paper 2009-3547) provides a direct view of the organization of the vortex lines from which the structural aspects of the flow may be understood. It is found that the common assumption that the physical form of coherent vortical structures should be synonymous with regions of rotational motion is unwarranted. In fact, what are normally referred to as hairpin vortices are found not to be vortical objects in their own right, but rather the rotational motion corresponding to vortical features of a very different kind. The latter may be described as raised streamwise-oriented furrows in the vortex layer overlying low speed streaks. Upstream, the uplifted vorticity has a streamwise projection producing counter-rotating motion normally described as being the legs of hairpin vortices. Downstream, the upwelling vorticity detaches from the wall acquiring a mushroom shape, with the illusion of hairpin legs continuing within counter-rotating motion produced by streamwise oriented filaments. Arch vortices are produced by roll-up of the shear layers along the top of the furrows. The existence of hairpin vortices is equivalent to imagining that regions of rotational motion contain vorticity forming a complete structure - in contradiction to the objects observed in the filament simulation. [Preview Abstract] |
Sunday, November 22, 2009 5:59PM - 6:12PM |
EA.00009: The dynamical role of vortex tubes and sheets in wall-bounded flows Sergio Pirozzoli Vortex sheets and tubes are extracted from DNS of a canonical compressible boundary layer, and their dynamical contribution analyzed by means of a non-local analysis based on the solution of the Poisson equation for the vector potential. The results show non-negligible contribution of vortex sheets to the wall layer dynamics, especially in the inner layer. The statistical relationship between tubes and sheets is also analyzed by means of conditional average fields extracted from a DNS database. The results support strong association between the two types of coherent structures, and indicate that vortex tubes are mainly produced upon roll-up of vortex sheets (as in the hairpin vortex paradigm), or interact causing the ejection of near-wall vorticity, or generate sheets of streamwise vorticity through a rubbing effect caused by the no-slip condition. [Preview Abstract] |
Sunday, November 22, 2009 6:12PM - 6:25PM |
EA.00010: Hairpin vortex trajectories in a turbulent boundary layer Gerrit Elsinga, Christian Poelma, Jerry Westerweel, Andreas Schroeder, Reinhard Geisler, Fulvio Scarano Hairpin vortices within a turbulent boundary layer have been tracked in space and time resulting in their average trajectory and spreading rates with respect to the average. This information is used to predict the space-time correlation function of different flow variables, such as the individual velocity components and the invariants of the velocity gradient tensor, given their respective spatial auto-correlation functions. The predictions are in good agreement with the actual space-time correlations over convection distances up to one boundary layer thickness, which indicates that the decreasing peak correlation values over time reflect the spreading rate rather than an actual topological evolution of the individual flow structures. The results will also be discussed in relation to Taylor's hypothesis. [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