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 R20: Boundary Layers IX: Numerical Simulation |
Hide Abstracts |
Chair: NR Panchapakesan, Indian Institute of Technology Madras Room: 315 |
Tuesday, November 26, 2013 1:05PM - 1:18PM |
R20.00001: Large eddy simulation study of the logarithmic law for high-order moments in turbulent boundary layers Richard Stevens, Michael Wilczek, Charles Meneveau Analyzing highly resolved experimental data, recently generalized log-laws have been observed by Meneveau and Marusic (JFM 719, R1 (2013)) for higher-order even moments of the streamwise velocity fluctuations $u'$. These observations provide new and robust data in canonical boundary layers that can be compared to the output of Large Eddy Simulation models. To utilize the new data for this purpose, we have performed large eddy simulations at various resolutions and with different sub-grid models (Smagorinsky and scale-dependent Lagrangian dynamic model). In agreement with the experimental data, the higher-order generalizations of the log-law are also observed in the LES results. But since large eddy simulations do not resolve the sub-grid contributions, we furthermore discuss the possibility of an a-posteriori correction of the statistical moments based on quasi-Gaussian assumptions for the statistics of the sub-grid contributions. We find that, while the corrections are small in the bulk of the flow, they can reach significant amplitudes in the near-wall region. [Preview Abstract] |
Tuesday, November 26, 2013 1:18PM - 1:31PM |
R20.00002: Inertial Subrange Spectra in the Log-Law Layer of Turbulent Channel Flow Yukio Kaneda, Koji Morishita, Takashi Ishihara High resolution direct numerical simulations (DNSs) of turbulent channel flows with the friction Reynolds number $Re_\tau$ up to 5120 show that there exists a layer at $y^+$ being approximately between 200 and 1200, in which the mean velocity profile and the diagonal components of the inertial subrange velocity correlation spectra fit well to the logarithmic law and the $k^ {-5/3} $ law, respectively. Here $y^+$ is the distance from the wall normalized by the wall unit, and $k$ is the wavenumber in the stream wise direction. The DNS data suggest that in the layer (log-law layer), there exists a high wave number range in which the influence of the mean flow on the turbulence statistics may be regarded to be small as compared to that of nonlinear interactions between the small-scale eddies of size $\sim 1/$(wave number), so that the former influence may be treated as a perturbation added to the turbulent state determined by the nonlinear turbulence dynamics in the absence of the mean flow. A perturbation analysis on the basis of this idea yields a simple prediction for the anisotropic velocity correlation spectra in the inertial subrange. The DNS data agree fairly well with the prediction. [Preview Abstract] |
Tuesday, November 26, 2013 1:31PM - 1:44PM |
R20.00003: Numerical experiments of thermal convection with shear Curtis Hamman, Parviz Moin Inspired by Sayadi, Hamman and Moin's (2013, J. Fluid Mech.) finding that late-stage boundary layer transition shares the same structure and scaling behaviour of high Reynolds number wall turbulence, we explore whether facets of the related ultimate state of Rayleigh-B\'{e}nard convection (RBC) can develop at moderate Rayleigh numbers ($10^5 < Ra < 10^{10}$) with amplified wall shear. In shear-free RBC, turbulence in the bulk produces a large-scale circulation or mean wind too weak to prompt boundary layer transition by shear instability, except possibly at extreme Rayleigh numbers where the largest eddies organize the local wall shear flow above a critical friction Reynolds number. We propose a numerical experiment to produce turbulent, three-dimensional, thermal and kinetic boundary layers near the walls while maintaining shear-free, buoyant turbulence production in the bulk as in unperturbed RBC. We speculate that this flow structure may correlate with the ultimate state of thermal convection, given that most of the near-wall turbulent energy production by shear is due to the effects of a mean wind-induced turbulent wall layer. Simulation results are presented to test this connection between the boundary layer structure of RBC and canonical wall-bounded turbulent shear flows. [Preview Abstract] |
Tuesday, November 26, 2013 1:44PM - 1:57PM |
R20.00004: LES of spatially developing turbulent boundary layer over a concave surface Sunil Arolla, Paul Durbin We revisit the problem of spatially developing turbulent boundary layer over a concave surface. Unlike previous investigations, we simulate the combined effects of curvature-induced pressure gradients as well as streamline curvature on the turbulence. Our focus is on investigating the response of the turbulent boundary layer to the sudden onset of curvature and the destabilizing influence of concave surface in the presence of pressure gradients. This is of interest for evaluating the turbulence closure models. Numerical simulations have been performed using the large eddy simulation framework in OpenFOAM. The dynamic Smagorinsky model is used to account for the sub-grid scale stresses. A variant of the recycling and rescaling method is used to generate the inflow turbulence. At the beginning of the curve, the momentum thickness Reynolds number is $1300$ and the ratio of boundary layer thickness to the radius of curvature is $\delta_{0}/R=0.055$. The radial profiles of the mean velocity and turbulence statistics at different locations along the concave surface are presented. In addition, the secondary flow structures observed are reported. [Preview Abstract] |
Tuesday, November 26, 2013 1:57PM - 2:10PM |
R20.00005: The high-order statistics of APG turbulent boundary layers Yvan Maciel, Ayse G. Gungor, Mark P. Simens, Julio Soria One and two-point statistics are presented from a new direct numerical simulation of an adverse pressure gradient boundary layer, at $Re_\theta=250-2175$, in which the transition to turbulence is triggered by a trip wire which is modeled using the immersed boundary method. Mean velocity results in the attached turbulent region do not show log law profiles. Departure from the law of the wall occurs throughout the inner region. The production and Reynolds stress peaks move to roughly the middle of the boundary layer. The profiles of the $uv$ correlation factor reveal that de-correlation between $u$ and $v$ takes place throughout the boundary layer, but especially near the wall, as the mean velocity defect increases. The non-dimensional stress ratios and quadrant analysis of $uv$ indicate changes to the turbulence structure. The structure parameter is low, similar to equilibrium APG flows and mixing layers in the present flow and seems to be decreasing as the mean velocity defect increases. The statistics of the upper half of the APG flow show resemblance with results for a mixing layer. [Preview Abstract] |
Tuesday, November 26, 2013 2:10PM - 2:23PM |
R20.00006: The structure of APG turbulent boundary layers Ayse G. Gungor, Yvan Maciel, Mark P. Simens, Julio Soria A boundary layer under influence of a strong APG is studied using DNS. Transition to turbulence is triggered using a trip wire which is modelled using the immersed boundary method. The Reynolds number close to the exit of the numerical domain is $Re_\theta=2175$ and the shape-factor $H=2.5$. Two dimensional two-point spatial correlation functions are obtained in this region and close to the transition region. $C_{vu}$ with a reference point close to the transition region shows a flow periodicity until $Re_{\theta} \approx 1600$. This periodicity is related to the shear layer instability of the separation bubble created as a result of the APG. The $C_{vv}$ and $C_{ww}$ correlations obtained far from the transition region at $Re_\theta=2175$ and at $y/\delta=0.4$ coincide with results obtained for a ZPG boundary layer. Implying that the structure of the $v,w$ fluctuations is the same as in ZPG. However, $C_{uu}$ indicates that the structure of the $u$ fluctuation in an APG boundary layer is almost twice as short as the ZPG structures. The APG structures are also less correlated with the flow at the wall. The near wall structure of strong APG flows is different from ZPG flows in that streaks are much shorter or absent. [Preview Abstract] |
Tuesday, November 26, 2013 2:23PM - 2:36PM |
R20.00007: Large eddy simulation of zero-pressure-gradient turbulent boundary layer based on different scaling laws Wan Cheng, Ravi Samtaney We present results of large eddy simulation (LES) for a smooth-wall, zero-pressure-gradient turbulent boundary layer. We employ the stretched vortex sub-grid-scale model in the simulations augmented by a wall model. Our wall model is based on the virtual-wall model introduced by Chung \& Pullin (J. Fluid Mech 2009). An essential component of their wall model is an ODE governing the local wall-normal velocity gradient obtained using inner-scaling ansatz. We test two variants of the wall model based on different similarity laws: one is based on a log-law and the other on a power-law. The specific form of the power law scaling utilized is that proposed by George \& Castillo (Appl. Mech. Rev. 1997), dubbed the ``GC Law''. Turbulent inflow conditions are generated by a recycling method, and applying scaling laws corresponding to the two variants of the wall model, and a uniform way to determine the inlet friction velocity. For Reynolds number based on momentum thickness, $Re_\theta$, ranging from $10^4$ to $10^{12}$ it is found that the velocity profiles generally follow the log law form rather than the power law. For large Reynolds number asymptotic behavior, LES based on different scaling laws the boundary layer thickness and turbulent intensities do not show much difference. [Preview Abstract] |
Tuesday, November 26, 2013 2:36PM - 2:49PM |
R20.00008: Turbulent Boundary Layers in Absence of Mean Shear Blair Johnson, Edwin Cowen Environmental flows are often observed in which turbulence levels significantly exceed what would be expected from mean boundary shear (e.g. breaking surface waves). This enhanced turbulence produces sediment resuspension and boundary layers that differ greatly from classic turbulent boundary layer characterizations. To identify the contribution of turbulence to such sediment resuspension, experiments are conducted in a facility designed to generate homogeneous isotropic turbulence in absence of mean shear via a Randomly Actuated Synthetic Jet Array (RASJA). Using particle image velocimetry (PIV), boundary layers above both a solid glass bed and a narrowly graded sediment bed are characterized by their mean flows, turbulent kinetic energy, dissipation, spectra, and Reynolds stress. Furthermore, a surprising observation includes the formation of ripple patterns when the turbulence decays above the sediment bed. We hypothesize that the ripples scale with the integral length scale of the turbulence. By varying the percentage of active jets and the relative on- and off-times of jets in the RASJA, our investigations consider the impact of altering the integral length scale of the facility on the resulting turbulent structures and sediment motions observed. [Preview Abstract] |
Tuesday, November 26, 2013 2:49PM - 3:02PM |
R20.00009: Investigation of Turbulent Wedge Spreading Mechanism and How to Reduce Spreading Using Surface Textures Jeff Chu, David Goldstein, Garry Brown We investigate the physics of turbulent wedge spreading in a nominally zero pressure gradient laminar boundary layer over a flat wall using incompressible spectral DNS and an immersed boundary method. Turbulent wedges are simulated over both physical and unphysical surfaces to identify the important factors leading to wedge spreading and turbulence regeneration. Vortex mechanics are examined in detail to elucidate the details of vorticity generation. We find that turbulent wedge spreading appears to rely on tilting spanwise vorticity into the streamwise direction. In particular, dw/dx appears to be an integral part of the tilting process. Low-speed streaks also appear to be important. We examine surface textures that could interfere with the spreading process including riblets and fins to interfere with the tilting process and dimples on the surface to lock in the spacing of low-speed streaks. [Preview Abstract] |
Tuesday, November 26, 2013 3:02PM - 3:15PM |
R20.00010: Similarities between statistically-stationary homogeneous shear turbulence and the logarithmic layer in channels Siwei Dong, Atsushi Sekimoto, Javier Jim\'enez The rough independence of the logarithmic layer (LL) of wall-bounded turbulence from the details of the buffer and outer layers, suggests that the interaction of the turbulent fluctuations with the mean shear may be mimicked by statistically-stationary homogeneous shear turbulence(SS-HST) in a finite box. We study SS-HST in boxes for which the statistics best agree with those of the LL. Both flows share similar Corrsin shear parameters, and Reynolds-stress and vorticity anisotropies. Two-point correlation functions show that $u$ and $w$ are constrained by the simulation box and are respectively shorter and narrower for SS-HST than for the LL, but $v$ and the vorticity are roughly of the same size in both flows when $Re_{\lambda}$ is similar. The transient bursting of $v$ in both flows is quite similar to the linear Orr mechanism, with time scales that are of the same order in both flows. In both cases, a streamwise velocity streak forms and breaks down quasi periodically, and the break down is accompanied by an enhanced flux of momentum, in the form of large-scale ejections and sweeps. [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. |
© 2022 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
1 Research Road, Ridge, NY 11961-2701
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700