Bulletin of the American Physical Society
63rd Annual Meeting of the APS Division of Fluid Dynamics
Volume 55, Number 16
Sunday–Tuesday, November 21–23, 2010; Long Beach, California
Session AB: Turbulent Boundary Layers I |
Hide Abstracts |
Chair: Joe Klewicki, University of New Hampshire/Melbourne Room: Long Beach Convention Center 101B |
Sunday, November 21, 2010 8:00AM - 8:13AM |
AB.00001: Extrapolating Channel Flow Data to High Reynolds Number Conditions Ricardo Vinuesa, Kris Dressler, Hassan Nagib Channel and pipe flows require relatively larger Reynolds numbers ($Re$) than boundary layers to achieve ``high-$Re$ conditions'' for wall-bounded flows. Unlike the pipe case, where the Superpipe results far exceed such conditions, even the highest Re channel experiments have not reached such a state. Because of the extremely large channel facility required to achieve these high-$Re$ conditions with good spatial resolution for the measurements, DNS may present the best hope. Available experimental and DNS results, and new data from a channel with variable aspect ratio, are used to estimate the high-$Re$ von K\'arm\'an coefficient and the $Re$ required to achieve it. Wall shear stress measurements with oil film interferometry were obtained over the range $7,500 < Re_m < 30,000$ in channel flows for aspect ratios varying from $18$ to $48$. The results show that the relationship between Reynolds number and skin-friction coefficient depends on the channel aspect ratio well beyond the traditional values of $8$ to $12$ generally believed to ensure two-dimensionality. The results indicate that an increase by a factor of about four or five in the current Re capabilities of DNS is required to approach the asymptotic von K\'arm\'an coefficient for the channel, which likely is even lower than our recent estimates of $0.37$. [Preview Abstract] |
Sunday, November 21, 2010 8:13AM - 8:26AM |
AB.00002: Modeling near-wall turbulent flows Ivan Marusic, Romain Mathis, Nicholas Hutchins The near-wall region of turbulent boundary layers is a crucial region for turbulence production, but it is also a region that becomes increasing difficult to access and make measurements in as the Reynolds number becomes very high. Consequently, it is desirable to model the turbulence in this region. Recent studies have shown that the classical description, with inner (wall) scaling alone, is insufficient to explain the behaviour of the streamwise turbulence intensities with increasing Reynolds number. Here we will review our recent near-wall model (Marusic~{\em et al.}, \emph{Science} 329, 2010), where the near-wall turbulence is predicted given information from only the large-scale signature at a single measurement point in the logarithmic layer, considerably far from the wall. The model is consistent with the Townsend attached eddy hypothesis in that the large-scale structures associated with the log-region are felt all the way down to the wall, but also includes a non-linear amplitude modulation effect of the large structures on the near-wall turbulence. Detailed predicted spectra across the entire near- wall region will be presented, together with other higher order statistics over a large range of Reynolds numbers varying from laboratory to atmospheric flows. [Preview Abstract] |
Sunday, November 21, 2010 8:26AM - 8:39AM |
AB.00003: Coherent vorticity extraction in turbulent boundary layers using orthogonal wavelets George Khujadze, Romain Nguyen van yen, Kai Schneider, Martin Oberlack, Marie Farge High resolution direct numerical simulation data of turbulent boundary layers are analyzed by means of wavelets. The developed anisotropic wavelet transform reinterpolates the data in the wall normal direction, originally given on a Chebychev grid, onto an adapted dyadic grid. The contructed wavelet bases accounts for the anisotropy of the flow by using different scales in the wall normal direction and in the planes parallel to the wall. Therewith the vorticity field is decomposed into coherent and incoherent contributions. It is shown that few wavelet coefficients retain the coherent structures of the flow, while the majority of the coefficients corresponds to a structurless noise like background flow. Scale and direction dependent statistics in wavelet space quantify the properties of the total, coherent and incoherent flows as a function of the wall distance. [Preview Abstract] |
Sunday, November 21, 2010 8:39AM - 8:52AM |
AB.00004: Large-eddy and direct simulations of accelerating boundary layers Junlin Yuan, Valerio Grazioso, Ugo Piomelli Turbulent boundary layers subject to a favorable pressure gradient (which induces freestream acceleration) are found in many engineering applications, such as airfoils or curved ducts. If the acceleration is sufficiently large, turbulence production decreases, and the flow reverts to a laminar or quasi-laminar state. Once the cause of relaminarization is removed, the flow re-transitions to turbulence in a process that may depend critically on the residual levels of turbulent fluctuation during the relaminarization. We performed direct and large-eddy simulations (DNS and LES) of accelerating boundary layers, on smooth and rough flat plates. The DNS allows to study both the relaminarization and re-transition without requiring any turbulence model that may alter the physics. It also validates the LES, which can be extended to higher Reynolds numbers. The roughness is included using an Immersed Boundary Method. The entrainment of the irrotational freestream fluid into the boundary layer plays a critical role in the formation of a well-mixed outer layer and the stabilization of the inner layer. The wall-normal and shear components of the Reynolds stress decay more rapidly than the streamwise one, leading to a state of inactive turbulence that is advected from the upstream boundary layer into the relaminarization region. Roughness effects are limited to the near wall, but are nonetheless visible. [Preview Abstract] |
Sunday, November 21, 2010 8:52AM - 9:05AM |
AB.00005: Direct Numerical Simulation of Zero-Pressure Gradient and Sink Flow Turbulent Boundary Layers O. Ramesh, Saurabh Patwardhan Direct Numerical Simulations have been performed for the zero pressure gradient (ZPG) (600 $<$ Re$_{\theta }<$ 900) and for the sink flow turbulent boundary layers (K = 7.71$\times $10$^{-7})$. A finite difference code on Cartesian grid was used to perform the simulations. Inflow generation method developed by Lund et al. was used to generate inflow boundary condition for the ZPG case. This method was slightly modified for the case of sink flow in view of self-similarity it possesses in the inner co-ordinates. Hence, there was no need to use empirical relations for the calculation of inlet $\theta $ or $\delta $ and rescaling in outer co-ordinates. The average statistics obtained from the simulations are in close agreement with the experimental as well as DNS data available in the literature. The intermittency distribution in the case of sink flow approaches zero inside the boundary layer (y = 0.8$\delta )$, an observation which is also confirmed by the experiments. This effect could be due to the acceleration near the boundary layer edge which suppresses the turbulent fluctuations near the boundary layer edge. [Preview Abstract] |
Sunday, November 21, 2010 9:05AM - 9:18AM |
AB.00006: Velocity-vorticity correlation structure in turbulent channel flow Jun Chen, Jie Pei, Zhen-Su She, Fazle Hussain A statistical structure -- velocity-vorticity correlation structure (VVCS) -- is defined by the amplitude distribution of a tensor field of correlation coefficicents. Applied to turbulent channel flow DNS database (at $Re_{\tau}=180$), it captures most relevant features -- qualitative and quantitative -- of coherent structures near the wall, including streaks (Kline et al. 1967, JFM), inclined streamwise vortices (Jeong et al. 1997, JFM), and transverse vorticity (Jimenez \& Moin 1991, JFM), etc. Associated with the streamwise velocity component (particularly $\langle u \omega_x \rangle$), VVCS reveals a change of topology with increasing $y_r^+$, providing a physical interpretation of multiple layers of wall-bounded turbulence. The statistical structure of $\langle u \omega_x \rangle$ depends on the $y_r^+$-location of $u$ detection. When $y_r^+$ is near the wall, the structure resembles streamwise vortices. But when $y_r^+$ is close to the center, it becomes a blob-like structure, quite different from streamwise vortices in the near-wall region. We propose that the statistical structure is adequate in modeling of the mean flow field. This study raises some doubt about unique structures in turbulent flows: consideration of a set of statistical structures is unavoidable. [Preview Abstract] |
Sunday, November 21, 2010 9:18AM - 9:31AM |
AB.00007: DNS of turbulent boundary layer over a flat plate at Re$_\theta=5200$ Antonino Ferrante, Keegan Webster We performed direct numerical simulations (DNS) of a spatially developing turbulent boundary layer over a flat plate at Re$_\theta=5200$. At this Reynolds number, our DNS results show that the overlap region of inner and outer layers extends for about $150$ wall units. The turbulent inflow conditions were generated using the method of Ferrante \& Elghobashi [J.~Comput.~Phys.~198 (2004)]. The computational domain of the main simulation is a parallelepiped with $2048\times1024\times512$ grid points in the streamwise, spanwise and wall-normal direction, respectively. The closest grid point to the wall is at $z^+=0.4$. The turbulence statistics were collected over a period of about 80 large-eddy turnover times. These simulations were made possible thanks to our development of an optimized and scalable 3D Poisson solver, which reduced the time to integrate the incompressible Navier-Stokes equations by 40\%. Our DNS results are in excellent agreement with the experimental data of DeGraaff and Eaton [J.~Fluid Mech.~422 (2000)] at the same Re$_\theta$. [Preview Abstract] |
Sunday, November 21, 2010 9:31AM - 9:44AM |
AB.00008: Direct Numerical Simulation of a Quasilaminarized Boundary Layer Luciano Castillo, Juan Guillermo Araya, Raul Bayoan Cal Direct Numerical Simulations of spatially-evolving turbulent boundary layers with strong favorable pressure gradients are performed. The driven force behind this investigation is elucidate the mechanisms responsible for the quasi-laminarization of the boundary layer. Budgets of the turbulent kinetic energy and the shear Reynolds stresses provide insight into the terms responsible for this phenomenon. The results also confirm the similarity analysis framework as develop by Cal and Castillo\footnote{R. B. Cal and L. Castillo (2008), Phys. Fluids. vol 20, 105106, 2008.} including the redistribution of the Reynolds stresses, a significant reduction in skin friction and a pressure parameter value which falls in the quasilaminar quadrant. The prescription of stronger favorable pressure gradients is mainly manifested by a significant decrease of the production of the shear Reynolds stresses and attenuation of the velocity-pressure gradient correlation term. The latter evidence confirms the important role of pressure fluctuations on the energy exchange and transport phenomena of flow parameters. [Preview Abstract] |
Sunday, November 21, 2010 9:44AM - 9:57AM |
AB.00009: DNS of Turbulent Boundary Layers under Highenthalpy Conditions Lian Duan, Pino Mart\'In To study real-gas effects and turbulence-chemistry interaction, direct numerical simulations (DNS) of hypersonic boundary layers are conducted under typical hypersonic conditions. We consider the boundary layer on a lifting-body consisting of a flat plate at an angle of attack, which flies at altitude $30$km with a Mach number $21$. Two different inclined angles, $35^{\circ}$ and $8^{\circ}$, are considered,representing blunt and slender bodies. Both noncatalytic and supercatalytic wall conditions are considered. The DNS data are studied to assess the validity of Morkovin's hypothesis, the strong Reynolds analogy, as well as the behaviors of turbulence structures under high-enthalpy conditions.Relative to low-enthalpy conditions [1], significant differences in typical scalings are observed. \\[4pt] [1] L. Duan and I. Beekman and M. P. Mart\'in, \emph{Direct numerical simulation of hypersonic turbulent boundary layers. Part 2: Effect of temperature}, J. Fluid Mech. {\bf 655} (2010), 419-445. [Preview Abstract] |
Sunday, November 21, 2010 9:57AM - 10:10AM |
AB.00010: Geometric study of Lagrangian and Eulerian structures in turbulent channel flow Yue Yang, Dale Pullin We report a geometric study of both evolving Lagrangian, and also instantaneous Eulerian structures in turbulent channel flow at $Re_\tau = 180,395, 590$. The former are obtained by tracking a Lagrangian scalar field while the latter are extracted from the swirling-strength field at a time instant. A multi-scale and multi-directional analysis, based on the mirror-extended curvelet transform, is developed to quantify flow structure geometry including the averaged inclination and sweep angles of both classes of flow structure at up to seven scales. These range from the half-height of the channel to several viscous length scales. Here, the inclination angle is defined on the plane of the stream-wise and wall-normal directions, and the sweep angle on the plane of the stream-wise and span-wise directions. Results for turbulent channel flow include the geometry of candidate hairpin vortices and other structures in the near-wall region, the structural evolution of near-wall vortices, and evidence for the existence and geometry of hairpin packets based on statistical inter-scale correlations. [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