Bulletin of the American Physical Society
61st Annual Meeting of the APS Division of Fluid Dynamics
Volume 53, Number 15
Sunday–Tuesday, November 23–25, 2008; San Antonio, Texas
Session BA: Turbulent Boundary Layers: Roughness Effects |
Hide Abstracts |
Chair: Xiaofeng Liu, Johns Hopkins University Room: 001A |
Sunday, November 23, 2008 10:30AM - 10:43AM |
BA.00001: Mean flow scaling in transitionally-rough turbulent boundary layers Michael P. Schultz, Karen A. Flack Results of an experimental investigation of the flow over several mildly-rough surfaces are presented. Three fine-grit sandpaper surfaces and two commercial ship bottom paints were tested over a large Reynolds number range (\textit{Re}$_{\theta}$ = 2,600 -- 30,000) in order to document the roughness function ($\Delta U^{+})$ behavior in the transitionally-rough flow regime. In all cases the root-mean-square roughness height was a very small fraction of the boundary layer thickness ($k_{rms}$/$\delta <$1/1,100). The results indicate that the mean velocity profiles for the rough surfaces agree with smooth-wall profiles using outer scaling. However, some significant differences in the behavior of $\Delta U^{+}$ in the transitionally-rough flow regime are noted among the five rough surfaces. For example, the roughness functions for the sandpaper surfaces show reasonable agreement with the results of Nikuradse for uniform sand, while the paint surfaces do not. These results, along with others from the literature, will be used to illustrate how surface topography may give rise to the differences that are observed in roughness functions for the transitionally-rough regime. [Preview Abstract] |
Sunday, November 23, 2008 10:43AM - 10:56AM |
BA.00002: Turbulence statistics for high Reynolds number rough wall turbulent boundary layers Karen A. Flack, Michael P. Schultz The effect of Reynolds number on rough wall turbulence statistics is investigated through a series of experiments on five rough surfaces. These consist of three fine-grit sandpaper surfaces and two commercial ship bottom paints, all with a roughness height that is a very small fraction of the boundary layer thickness. Reynolds number dependence of the near wall peak in the streamwise Reynolds normal stress is well established for smooth walls, and it has been observed that this peak is destroyed in the fully rough turbulent boundary layer. Additionally, Reynolds number effects have been observed in the streamwise Reynolds stress in the log-law and outer layers for smooth walls. This Reynolds number dependence has not been tested over a wide Reynolds number range for rough walls. Results spanning a Reynolds number range of \textit{Re}$_{\theta }$ = 3,000 - 30,000 will be presented. [Preview Abstract] |
Sunday, November 23, 2008 10:56AM - 11:09AM |
BA.00003: Diffusion of Drag-Reducing Polymers within a High-Reynolds-Number, Rough-Wall Turbulent Boundary Layer Brian Elbing, Marc Perlin, David Dowling, Michael Solomon, Steven Ceccio Two experiments were conducted to investigate polymer drag reduction (PDR) within high Reynolds number (to 200 million based on downstream distance), rough-wall turbulent boundary layers. The first experiment was conducted at the U.S. Navy's Large Cavitation Channel on a 12.9 m long flat-plate at speeds to 20 m/s with the surface hydraulically smooth and fully rough. Local skin-friction measurements on the smooth and rough surfaces had maximum PDR levels of 65 and 75 percent, respectively. However, PDR decreased with increasing downstream distance and flow speed more rapidly on the rough surface, and at the top speed no measureable level of PDR was observed. The roughness-induced increased diffusion was quantified with near-wall concentration measurements and the second experiment, which measured concentration profiles on a 0.94 m long flat-plate with three surface conditions: smooth, 240-grit, and 60-grit sandpaper. The increased diffusion does not fully explain the smooth-rough PDR differences observed in the first experiment. Rheological analysis of drawn samples from the first experiment indicates that polymer degradation (chain scission) could be responsible for the remaining loss of rough-wall PDR. These results have implications for the cost effectiveness of PDR for surface ships. [Preview Abstract] |
Sunday, November 23, 2008 11:09AM - 11:22AM |
BA.00004: Large Eddy Simulation of Turbulent Boundary Layers over Rough Beds Krishnakumar Rajagopalan, Geno Pawlak, Antoine Patalano, Marcelo Kobayashi The hydrodynamic length scale, $z_{o},$ a parameter in the logarithmic velocity profile for fully developed turbulent boundary layers over rough beds, is a simple function of the physical size of the roughness elements when the roughness is homogeneous. In the case of broad-banded and highly irregular roughness distributions as found, for example, over coral reefs, the relationship between the hydrodynamic length scale and the physical length scale is not clear. One method to characterize the irregular nature of roughness is through a spectral distribution (Nunes and Pawlak, \textit{J.Coast.Res}, 2008). We present results from numerical and laboratory experiments that model spectral roughness using inhomogeneous distributions of square waves which can reproduce flow separation characteristics common to flow over rough beds. The numerical simulations make use of Large Eddy Simulation techniques to model the turbulent flow over the bed. Laboratory observations are presented to verify the numerical results. This study will shed light on estimating bulk parameters of the flow such as hydrodynamic length scale, mean velocity profile and bed stress for a boundary with a spectral roughness distribution. [Preview Abstract] |
Sunday, November 23, 2008 11:22AM - 11:35AM |
BA.00005: Refined Analysis of the Mean Momentum Balance in Rough-Wall Turbulent Boundary Layers Faraz Mehdi, Joseph Klewicki, Christopher White Determining the dominant terms in the mean momentum balance (MMB) as a function of position within a turbulent flow reveals the important dynamical mechanisms, and can form the basis for multiscale analysis. Little is known regarding the combined influences of roughness and Reynolds number on behavior of the MMB in turbulent boundary layers. Previous analyses reveal that this is especially the case in and near the roughness sublayer where the data scatter is significant. A semi-theoretical method for reducing the effect of this scatter on the empirical determination of the MMB is described. The efficacy of this method is demonstrated via its capacity to recover the existence of a near-wall stress gradient balance layer that by simple physical considerations is known to exist. Refined MMB analyses augment previous preliminary results based on the location of the Reynolds stress peaks for different $\delta^+$ and $k_s^+$. Contrary to prevalent notions regarding rough-wall layers, the present results strongly suggest the importance of the viscous force in, and in some cases, above the roughness sublayer, and that relative to the establishment of the mean momentum field there is a significant coupling between the combined effects of roughness and Reynolds number. The support of the NSF (CTS0555223, grant monitor William Schultz) and the ONR (N000140810836, grant monitor Ronald Joslin) is gratefully acknowledged. [Preview Abstract] |
Sunday, November 23, 2008 11:35AM - 11:48AM |
BA.00006: Comparison of boundary layers over 2-D and 3-D rough walls Ralph J. Volino, Michael P. Schultz, Karen A. Flack A zero pressure gradient boundary layer over a surface with two- dimensional, k-type roughness was investigated experimentally. The roughness consisted of transverse bars of square cross section spaced 8 bar heights apart. Previous work on a wide variety of surfaces with three-dimensional roughness has shown very good similarity between rough and smooth wall boundary layers outside the roughness sublayer. Similarity was observed in Reynolds stresses, higher order moments, swirl strength and two point correlations. The roughness sublayer thickness was considered to be about 3 equivalent sand grain roughness heights, $k_s$, and $k_s$ was always of the same order of magnitude as the roughness element height. On the 2-D bars, $k_s$ was an order of magnitude larger than the bar height, encompassing most of the boundary layer thickness. Reynolds stresses were noticeably larger over the 2-D roughness than the 3-D roughness, and length scales based on two-point spatial correlations were longer in all directions for all quantities with the 2-D roughness. The differences between 2-D and 3-D roughness observed in boundary layers have not been observed in channel flow. Reasons for the differences will be discussed. [Preview Abstract] |
Sunday, November 23, 2008 11:48AM - 12:01PM |
BA.00007: Velocity scaling in very-rough-wall channel flows David Birch, Jonathan Morrison Fully-developed turbulent channel flow over mesh-type and grit- type surface roughness topologies is compared at similar roughness Reynolds numbers $ku_{\tau}/{\nu}$ (where $k$ is the roughness height) in order to investigate the effect of the specific roughness geometries upon outer layer similarity for cases where $k/h > 3\%$, where $h$ is the channel half- height. Outer-scaled mean velocity profiles are found to collapse for $y/h > 0.15$, while for $y/h < 0.15$ the flow is subject to local inhomogeneities stemming from the individual roughness element wakes. Outer-scaled self-similarity is observed in the second and fourth velocity moments for $y/h > 0.2$, while the third moment collapses only for $y/h > 0.4$. With the inner length-scale defined as $(y-d)/y_0$ (where $d$ is the zero-plane offset and $y_{0}$ is the roughness length scale, found to be $0.0026h$ and $0.0085h$ for the grit and mesh surfaces, respectively), the mean velocity over the grit surface exhibits inner-scaled similarity for $(y-d)/h < 0.06$, while no region of inner-scaled self-similarity was apparent over the mesh surface. [Preview Abstract] |
Sunday, November 23, 2008 12:01PM - 12:14PM |
BA.00008: Zero Pressure Gradient Turbulent Boundary Layer Response to a Spatially Impulsive Dynamic Roughness Jeffrey LeHew, Ian Jacobi, Beverley McKeon The effect of a surface with a temporally actuated, spatially impulsive patch of roughness on a zero pressure gradient turbulent boundary layer is experimentally studied. This study is of potential interest for flow control applications and for understanding the effect that this additional roughness time scale has on turbulent boundary layers, specifically how it influences the relaxation of a turbulent boundary layer back to equilibrium after encountering the roughness. The time varying surface topology of the roughness is characterized using a scanning laser displacement sensor with phase locking capabilities. Boundary layer profiles are taken at downstream locations for a range of roughness excitation frequencies and amplitudes. Results are compared to smooth wall measurements as well as measurements made over similarly shaped static roughness elements. [Preview Abstract] |
Sunday, November 23, 2008 12:14PM - 12:27PM |
BA.00009: LES analysis of roughness scale effect on rough-wall turbulent boundary layers Kojiro Nozawa, Tetsuro Tamura Large Eddy Simulation (LES) of turbulent boundary layer flows over small-scale and large-scale homogenous roughness were performed. The turbulent boundary layer flow over small-scale roughness whose roughness height is 0.028$\delta$ (where $\delta$ is boundary layer thickness) is expected to have outer-layer similarity in the turbulence structure as same as the case of smooth wall. While the height of large-scale roughness is so large (0.14$\delta$) that roughness effects on the turbulence extend across the entire boundary layer, and the concept of wall similarity will be invalid [M. P. Schultz and K. A. Flack, J. Fluid Mech., 580, 381--405, (2007)]. In this study, in order to realize the spatially developing turbulent boundary layer with no pressure gradient, the quasi-periodic boundary condition is applied in streamwise direction, in which the velocities at the recycle station are rescaled and reintroduced at the inlet and the outflow boundary is set far downstream of the recycle station. We focused on the vertical distribution of second invariant of velocity gradient tensor ($Q$) which could identify vortical structures formed in the wake of roughness blocks with various scales, and showed physical understanding of turbulent flow structures based on its resulting distributions which have changed the range of log-law region. [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