Bulletin of the American Physical Society
66th Annual Meeting of the APS Division of Fluid Dynamics
Volume 58, Number 18
Sunday–Tuesday, November 24–26, 2013; Pittsburgh, Pennsylvania
Session L20: Boundary Layers VI: Channel Flow and Flows over Superhydrophobic Walls |
Hide Abstracts |
Chair: R. Rubinstein, NASA Langley Room: 315 |
Monday, November 25, 2013 3:35PM - 3:48PM |
L20.00001: Representation of the velocity spectra and Reynolds stress co-spectrum in turbulent channel flow using resolvent modes Rashad Moarref, Ati S. Sharma, Joel A. Tropp, Beverley J. McKeon We represent the velocity field in channel flow as a weighted sum of a small number of `resolvent modes' that are obtained by Fourier decomposition in the wall-parallel directions and time, and singular value decomposition of the resolvent operator in the wall-normal direction, following McKeon {\&} Sharma (J. Fluid Mech., 2010). Building on previous efforts in which the Reynolds number scaling and geometric self-similarity of the resolvent modes were identified in a study of the streamwise velocity variance, we determine the resolvent mode weights required to minimize the deviation between an assembly of resolvent modes at Re\textunderscore tau $=$ 2003 and the time-averaged two-dimensional spectra (uu, vv, ww and uv) from direct numerical simulations (Hoyas {\&} Jimenez, Phys. Fluids, 2006). While the spectra corresponding to small wavelengths can be approximated by a few resolvent modes, a larger number of modes is necessary for matching at large wavelengths. The Reynolds number scaling of the spectra and the associated implications of previously-identified self-similar attached eddies are further discussed. [Preview Abstract] |
Monday, November 25, 2013 3:48PM - 4:01PM |
L20.00002: High-Reynolds-number effects in turbulent channel flow: evidence from DNS Matteo Bernardini, Sergio Pirozzoli, Paolo Orlandi The behavior of the incompressible turbulent channel flow is investigated through direct numerical simulation up to a Reynolds number ($Re_{\tau} \approx 4080$) at which phenomena typical of the asymptotic Reynolds number regime starts to be observed. Less than a decade of nearly-logarithmic variation is observed in the mean velocity profiles, with log-law constants $k \approx 0.386$, $C \approx 4.30$. A log layer is also observed in the spanwise velocity variance, as predicted by Townsend's attached eddy hypothesis, whereas the streamwise variance seems to exhibit a plateau, perhaps being still affected by low-Reynolds-number effects. Comparison with previous DNS data at lower Reynolds number suggests strong enhancement of the imprinting effect of outer-layer eddies onto the near-wall region. This mechanisms is shown to be associated with excess turbulence kinetic energy production in the outer layer, and it clearly reflects in flow visualizations and in the streamwise velocity spectra, which exhibit near-tonal behavior in the outer layer. Associated with the outer energy production site, we find evidence of a Kolmogorov-like inertial range, limited to the spanwise spectral density of $u$, whereas power laws with different exponents are found for the other spectra. [Preview Abstract] |
Monday, November 25, 2013 4:01PM - 4:14PM |
L20.00003: Turbulent spots in a channel: an experimental study on the inner structure and the large-scale flow Jos\'e Eduardo Wesfreid, Gr\'egoire Lemoult, Konrad Gumowski, Mingyang Luo, Jean Luc Aider Transition to turbulence in plane Poiseuille flow in channels, occurs in presence of localized coherent structures, known as turbulent spots, composed of an assemblage of small-scale longitudinal vortices. We present an exhaustive experimental description of these spots, in a water channel of rectangular cross section. The test section's half height is h $=$ 10mm, its length is 220h, and its width is 15h. We study the response of the flow to a single, short perturbation, which above Re $=$1300 will always trigger the development of a turbulent spot. The fine structure of the flow field inside and around a turbulent spot is obtained from the measurements of the three components of the velocity fields~in a~cross sectionnal plane with Time Resolved Stereoscopic Particule Image Velocimetry. The flow in the turbulent can be decomposed into a large-scale motion consisting of an asymmetric quadrupole centred on the spot and a small-scale part consisting of streamwise streaks. From the observations of the temporal evolution of the energy of the streamwise and spanwise velocity perturbations, it is suggested that a self-sustaining process can occur in a turbulent spot above a given Reynolds number. We also compare the dynamical evolution of the turbulent fluctuations and the mean flow distortions, with the predictions of reduced models predicting the main features of subcritical transition to turbulence. [Preview Abstract] |
Monday, November 25, 2013 4:14PM - 4:27PM |
L20.00004: Temperature fluctuations in fully-developed turbulent channel flow with heated upper wall Carla Bahri, Michael Mueller, Marcus Hultmark The interactions and scaling differences between the velocity field and temperature field in a wall-bounded turbulent flow are investigated. In particular, a fully developed turbulent channel flow perturbed by a step change in the wall temperature is considered with a focus on the details of the developing thermal boundary layer. For this specific study, temperature acts as a passive scalar, having no dynamical effect on the flow. A combination of experimental investigation and direct numerical simulation (DNS) is presented. Velocity and temperature data are acquired with high accuracy where, the flow is allowed to reach a fully-developed state before encountering a heated upper wall at constant temperature. The experimental data is compared with DNS data where simulations of the same configuration are conducted. [Preview Abstract] |
Monday, November 25, 2013 4:27PM - 4:40PM |
L20.00005: Reynolds Number Effects on Kinetic Energy Transfer from Outer Layer in Turbulent Channel Flows Yoshinobu Yamamoto, Yoshiyuki Tsuji Kinetic energy transfer from the outer layer to the inner layer was investigated by means of DNS database of turbulent channel flows up to Re$\tau = $ 2000. On assumption of Dean's empirical equation, energy transfer from the mean velocity component to the turbulence was increased in proportion of 1/7 power of turbulent Reynolds number, but the peak value of the turbulent production was saturated with Re$\tau = $ 6000. This might be supported the importance of the second peaks of turbulent intensities in high-Re. In the view points of the ratio of kinetic energy transfer, both results of DNS database and theoretical analysis using logarithmic mean velocity profile represent that the boundary between the outer layer and the inner layer will be determined as the 20{\%} of the channel half-length. [Preview Abstract] |
Monday, November 25, 2013 4:40PM - 4:53PM |
L20.00006: An investigation of the flow modification in a turbulent channel with gain-based optimal forcing Arjun Sharma, Rashad Moarref, Mitul Luhar, David Goldstein, Beverley McKeon Direct numerical simulations of turbulent channel at Re$_{\mathrm{\tau \thinspace }}=$ 180 with two-dimensional traveling wave forcing are performed. The chosen forcing parameters, that is, the wall-parallel wavelength and wave-speed, are representative of the near-wall cycle. The wall-normal forcing shape is obtained using the resolvent analysis of McKeon {\&} Sharma (Journal of Fluid Mechanics, vol. 658, 2010). The results obtained from direct simulations are found to compare well with the resolvent analysis predictions for different forcing amplitudes. The turbulence statistics and visualizations of three-dimensional unsteady flow field shed light on the flow evolution in response to forcing. [Preview Abstract] |
Monday, November 25, 2013 4:53PM - 5:06PM |
L20.00007: Skin-friction Drag Reduction in Turbulent Channel Flow with Idealized Superhydrophobic Walls Amirreza Ratsegari, Rayhaneh Akhavan Skin-friction drag reduction by super-hydrophobic (SH) surfaces was investigated using Lattice Boltzmann DNS in turbulent channel flow with SH longitudinal microgrooves on both walls. The liquid/gas interfaces in the SH microgrooves were modeled as flat, shear-free surfaces. Drag reductions (DR) ranging from $5\%$ to $47\%$ were observed for microgrooves of size $4 \le g^{+0} = w^{+0} \le 128$ in channels of bulk Reynolds number $Re_b= U_b h/\nu = 3600$ ($Re_{\tau_0} = u_{\tau_0}h/\nu \approx 230$), where $g^{+0}$ and $w^{+0}$ denote the widths of the slip and no-slip surfaces in base flow wall units. It is shown that in both laminar and turbulent flow, DR scales as $DR = U_s/U_b + \varepsilon$. In laminar flow, where DR is purely due to surface slip, $\varepsilon=0$. In turbulent flow, $\varepsilon$ remains negligible when the slip length is smaller than the thickness of the viscous sublayer. For $DR>40\%$, where the effect of surface slip can be felt in the buffer layer, $\varepsilon$ attains a small non-zero value. Analysis of turbulence statistics and turbulence kinetic energy budgets confirms that outside of a layer of size approximately one slip length from the walls, the turbulence dynamics proceeds as in regular channel flow with no-slip walls. [Preview Abstract] |
Monday, November 25, 2013 5:06PM - 5:19PM |
L20.00008: A Chracterization of Superhydrophobic Surfaces for Skin-Friction Drag Reduction Hyunwook Park, John Kim A proper characterization of superhydrophobic surfaces (SHSs) was examined using direct numerical simulation of turbulent channel flows. A SHS was modeled through a shear-free boundary condition on the air-water interface. Within the considered Reynolds number range and SHS geometry, it was found that the drag reduction in turbulent channel flows was well correlated with the effective slip length normalized by viscous wall units. A maximum drag reduction was achieved when the effective slip length was on the order of 50 in viscous wall units. It was also shown that near-wall turbulence structures were significantly modified. Since the effective surface slip length can be interpreted as a depth of influence into which SHS affect the flow in the wall-normal direction, this result demonstrated that a SHS achieved its drag reduction by affecting those turbulence structures within the buffer layer of the turbulent channel flow. The present results also showed that the relative size between near-wall turbulence structures and the SHS geometry was an important parameter for drag reduction. [Preview Abstract] |
Monday, November 25, 2013 5:19PM - 5:32PM |
L20.00009: Direct numerical simulation of turbulent flows over superhydrophobic surfaces: gas-liquid interface dynamics Jongmin Seo, Ricardo Garc\'Ia-Mayoral, Ali Mani Superhydrophobic surfaces can induce large slip velocities for liquid flows, reducing the skin friction on walls, by entrapping gas pockets within the surface roughness. This work explores the onset mechanism leading to gas depletion through interface breakage under turbulent conditions. We conduct direct numerical simulations of flows over superhydrophobic walls. The superhydrophobic texture is conventionally modeled as a pattern of slip/no-slip boundary conditions for the wall-parallel velocities but, to take into account the dynamic deformation of the gas-liquid interface, we also introduce non-zero boundary conditions for the wall-normal velocity. These conditions are derived from the deformation of the interface in response to the overlying turbulent pressure fluctuations, following the Young-Laplace equation. Surface protrusions in the form of posts and streamwise-aligned ridges are studied, and results are presented as a function of the ``deformability'' of the gas-liquid interfaces, expressed as a Weber number. We will also discuss results for misaligned ridges. [Preview Abstract] |
Monday, November 25, 2013 5:32PM - 5:45PM |
L20.00010: Convective Air Mass Transfer in Submerged Superhydrophobic Surfaces:\ Turbulent Flow Christina A. Barth, Hooman Vahedi Tafreshi, Mohamed Gad-el-Hak Longevity of entrapped air is an outstanding problem for using superhydrophobic coatings. Herein, we analyze from first principles a mass transfer problem. Using integral methods, we are able to extend our laminar flow solution presented last year to turbulent flows. We introduce an effective slip to the hydrodynamic boundary layer using a modified 1/7-power law velocity profile. We then introduce the hydrodynamic solution to the two-dimensional problem of alternating solid--water and air--water interfaces to determine the convective mass transfer of air's dissolution into water. This situation simulates spanwise microridges. The decoupled mass-transfer problem is solvable using an approximate integral method previously optimized by Reynolds et al.\ (1958). A mass-transfer correlation is derived as a function of the surface geometry (or gas area fraction), Reynolds number, and Schmidt number. Longevity, or time-dependent hydrophobicity, can be estimated from the resulting mass-transfer correlation. As expected, turbulence greatly enhances the rate of convective mass transfer, and thus superhydrophobicity is not maintained as long as it would under corresponding laminar flow conditions. [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