Bulletin of the American Physical Society
64th Annual Meeting of the APS Division of Fluid Dynamics
Volume 56, Number 18
Sunday–Tuesday, November 20–22, 2011; Baltimore, Maryland
Session G7: Turbulent Boundary Layers IV |
Hide Abstracts |
Chair: Beverly McKeon, California Institute of Technology Room: 310 |
Monday, November 21, 2011 8:00AM - 8:13AM |
G7.00001: Artificial large-scale-motion perturbation of a turbulent boundary layer I. Jacobi, B.J. McKeon A zero-pressure-gradient flat-plate boundary layer is perturbed dynamically with a spatially-impulsive strip of two-dimensional roughness elements (which alternate with a flush-surface condition periodically in time). The perturbation knocks the boundary layer out of equilibrium and artificially introduces a very-large-scale, periodic structure into the flow. Large-scale- motions in the turbulent boundary layer have recently been understood as a significant source of both turbulent kinetic energy and Reynolds stress, and in addition, have been shown to take part in an apparent amplitude modulation of smaller scales in the flow. The properties of this artificially generated large- scale-motion are studied with particular emphasis on the phase relationship between it and smaller scale structures. The behavior of the artificially-introduced large-scale-motion is also compared with the natural large-scales of the unperturbed turbulent boundary layer, and the interaction between the artificial and natural large-scales is explored. This study is supported by the Air Force Office of Scientific Research under grant \#FA9550-08-1-0049 (program manager John Schmisseur). [Preview Abstract] |
Monday, November 21, 2011 8:13AM - 8:26AM |
G7.00002: What happens to wall-bounded flows when the mean velocity profile is artificially prescribed? Javier Jimenez, Florian Tuerke Direct numerical simulations of turbulent channels with artificially prescribed mean velocity profiles, both natural and purposely unnatural, are used to study the dynamics of the energy-containing turbulent fluctuations. It is found that turbulence develops correctly in natural profiles, but that even slightly incorrect mean velocity gradients modify the intensities and Reynolds stresses substantially. The extra energy created by a locally stronger imposed shear resides in structures with essentially correct dimensionless ratios, but which are out of energy equilibrium. In profiles with sharp shear jumps in the logarithmic layer, the relaxation rates away from the discontinuity are different for the kinetic energy, the tangential Reynolds stress, and the dissipation, but they are all consistent with the turbulent advection of eddies with relaxation times of the order of the local eddy turnover. Correspondingly, the energy imbalance is compensated by turbulent diffusion. Interestingly, only eddies above a certain size are created by the discontinuity. The smallest ones do not immediately attach to the wall, but those about twice as large do. [Preview Abstract] |
Monday, November 21, 2011 8:26AM - 8:39AM |
G7.00003: Comparison of the one- and two-point second-order statistics for zero-pressure-gradient turbulent boundary layers and channels at $Re_\tau \approx 1000-2000$ Juan A. Sillero, Ayse G. Gungor, Javier Jim\'enez, Robert D. Moser One and two-points statistics are presented from a new direct simulation of the zero-pressure-gradient turbulent boundary layer in the range $Re_\theta = 2780-6680$, matching channel flow simulations at $Re_\tau \approx 1000-2000$. The integral parameters, mean velocities, Reynolds stresses and pressure fluctuations of the boundary layer closely agree with the numerical and experimental data sets available in the literature, but show clear differences with channels, when expressed in wall units at the same $y/\delta$. Those differences seem to increase, rather than decrease, with the Reynolds number. Spectra and spatial correlations $C_{\xi\xi}(x; x'; y; y'; k_z)$ are also investigated, and include a range of scales of the energy-containing structures larger than an order of magnitude. [Preview Abstract] |
Monday, November 21, 2011 8:39AM - 8:52AM |
G7.00004: Effect of the outer laminar flow on zero-pressure-gradient turbulent boundary layers Yoshinori Mizuno, Callum Atkinson, Julio Soria This study examines the effect of the invasion of the laminar flow on zero-pressure-gradient turbulent boundary layers by using direct numerical simulations, comparing with channel flows for comparable Reynolds numbers around $\delta^+=600$, where $\delta$ is the boundary layer thickness and ${}^+$ stands for normalization by a wall-unit. The turbulent/non-turbulent boundary is captured in an appropriate way, and the transitional layer, often called viscous (or laminar) superlayer, is defined. The distribution of the position of the boundary is represented well by the Gaussian distribution function of the wall-distance, and the mean and standard deviation are found to be approximately $0.9\delta$ and $0.15\delta$, respectively. A clear gap in turbulence intensities is observed within this layer, and enstrophy and dissipation rate are locally enhanced by the strong shear in the turbulence-side of this layer. The thickness of the viscous superlayer is found to be in the order of the Kolmogorov's length as found in previous works, but it is also found that it becomes thinner as moving away from the wall. The enhancement of enstrophy and dissipation rate is more significant at higher positions. [Preview Abstract] |
Monday, November 21, 2011 8:52AM - 9:05AM |
G7.00005: Development of a turbulent boundary layer beneath finite-amplitude continuous freestream turbulence Xiaohua Wu, Parviz Moin Following the earlier work of Wu {\&} Moin (JFM 2009, PoF 2010) and Wu (JFM 2010), here we will present our third, most recent, direct numerical simulation of the incompressible, zero-pressure-gradient flat-plate boundary layer. Heat transfer between the constant-temperature plate and the free-stream is also simulated with unit molecular Prandtl number. The freestream of the present boundary layer has continuous isotropic turbulence whose inlet strength is 3{\%} of the mean velocity. Its decay characteristics agree with existing water channel experiments. Despite the finite-level freestream perturbation, the boundary layer is clean in the sense that the deviation of skin-friction from Blasius prior to breakdown is less than 1{\%}. Both the statistics and structures from this simulation will be compared with our previous DNS studies using periodically fed patches of isotropic turbulence. The associated bypass transition process will also be evaluated. [Preview Abstract] |
Monday, November 21, 2011 9:05AM - 9:18AM |
G7.00006: Direct numerical simulation of K-type and H-type transitions to turbulence in a low Mach number flat plate boundary layer Taraneh Sayadi, Curtis Hamman, Parviz Moin Transition to turbulence via spatially evolving secondary instabilities in compressible, zero-pressure-gradient flat plate boundary layers is numerically simulated for both the Klebanoff K-type and Herbert H-type disturbances. The objective of this work is to evaluate the universality of the breakdown process between different routes through transition in wall-bounded shear flows. Each localized linear disturbance is amplified through weak non-linear instability that grows into lambda-vortices and then hairpin-shaped eddies with harmonic wavelength, which become less organized in the late-transitional regime once a fully populated spanwise turbulent energy spectrum is established. For the H-type transition, the computational domain extends from $Re_x = 10^5$, where laminar blowing and suction excites the most unstable fundamental and a pair of oblique waves, to fully turbulent stage at $Re_x = 10.6\times10^5$. The computational domain for the K-type transition extends to $Re_x = 9.6\times 10^5$. The computational algorithm employs fourth-order central differences with non-reflective numerical sponges along the external boundaries. For each case, the Mach number is 0.2. [Preview Abstract] |
Monday, November 21, 2011 9:18AM - 9:31AM |
G7.00007: DNS of laminar/turbulent boundary layer transition induced by solid obstacles Paolo Orlandi, Matteo Bernardini, Sergio Pirozzoli Direct numerical simulation is used to investigate how boundary layer transition is affected by the shape and size of an isolated obstacle whose size is of the order of the boundary layer thickness. The Navier-Stokes equations are discretized by means of an energy-conserving second-order staggered finite-difference method, and the geometrical complexity associated with the obstacle is handled through the immersed-boundary technique. A series of simulations have been performed by varying: i) the obstacle shape (cylinders and prisms with rectangular and triangular base); ii) the roughness height (as a fraction of the boundary layer thickness); iii) the width of the obstacle; iv) the Reynolds number of the incoming boundary layer. We have monitored the vorticity dynamics of the structures which are shed past the obstacle, and observed the concurrence of two mechanisms which promote transition to turbulence, namely the unsteady shear layer separation at the top edge of the obstacle, and the regeneration of quasi-streamwise vortices at the sides of the obstacle. The validity of semi-empirical transition criteria based on a suitably defined roughness Reynolds number will also be discussed, and associated with the physical mechanisms responsible for the self-sustainment of the disturbances. [Preview Abstract] |
Monday, November 21, 2011 9:31AM - 9:44AM |
G7.00008: Turbulent boundary layers: Inflow effects and cross-validation of simulation and experiment Ramis Oerlue, Philipp Schlatter A recent assessment of available direct numerical simulation (DNS) data from turbulent boundary layer flows [Schlatter \& \"Orl\"u, J. Fluid Mech.\ 659, 116 (2010)] showed surprisingly large differences not only in the skin friction coefficient or shape factor, but also in their predictions of mean and fluctuation profiles far into the sublayer. Several DNS of a zero pressure-gradient (ZPG) turbulent boundary layer (TBL) \`a la Schlatter et al.~[Phys. Fluids 21, 051702 (2009)] with physically different inflow conditions and tripping effects were performed. Most of the differences observed when comparing available DNS could thereby be traced back to different initial conditions. It was also found, that if transition is initiated at a low enough Reynolds number (based on the momentum-loss thickness) $Re_\theta<$~300, all data agree well for both inner and outer layer for $Re_\theta>2000$; a result that gives a lower limit for meaningful comparisons between numerical and/or wind tunnel experiments. Based on these results a detailed comparison between DNS and experiment of a ZPG TBL flow at $Re_\theta=2500$ and 4000 is presented. Good agreement is obtained for integral quantities, mean and fluctuating streamwise velocity profiles, but also for the probability distribution and spectral map throughout the boundary layer. [Preview Abstract] |
Monday, November 21, 2011 9:44AM - 9:57AM |
G7.00009: Turbulent boundary layers in long computational domains Philipp Schlatter, Qiang Li, Ramis Oerlue, Geert Brethouwer, Arne V. Johansson, P. Henrik Alfredsson, Dan S. Henningson A new series of numerical simulations of spatially evolving turbulent boundary layers is discussed. The very long computational domain starts at a low $Re_\theta=180$, where laminar-turbulent transition is initiated, reaching up to the (computationally very) high $Re_\theta=8500$. In the domain, the boundary layer develops naturally from the tripping location to the higher Reynolds numbers without any re-injection or recycling procedures. In consequence, this computational setup allows us to study, e.g., the mean flow development and the scaling behavior of the fluctuating energy free from pseudo-periodic effects. However, such domains require a large number of grid points; in the present case up to 10 billion for running well-resolved large-eddy simulation. The present results show excellent agreement with wind-tunnel experiments at similar $Re$ and previous (lower-$Re$) simulations (both direct and large- eddy simulations). The mean velocity profiles closely follow the correlation proposed by Monkewitz et al. (2007), just about reaching the plateau in the log-law diagnostic function. In a second part, three-dimensional visualizations of the evolving turbulent boundary layer are discussed with special focus on the persistence of transitional flow structures towards higher Reynolds numbers, having a highly unordered appearance. [Preview Abstract] |
Monday, November 21, 2011 9:57AM - 10:10AM |
G7.00010: Turbulence statistics in turbulent spots in a transitional boundary layer subject to free-stream turbulence Brendan Rehill Within the boundary layer transition region turbulent spots emerge and grow to form the fully-turbulent boundary layer. This paper examines the turbulent statistics within turbulent spots in a transitional boundary layer subject to free-stream turbulence intensity of $4.7\%$. Conditionally sampled DNS results, where the laminar and turbulent contributions to the transition region are separated, are used to obtain the relevant statistics. Conditional sampling of the data provides some improvement over the more classical time-space-averaged data reduction techniques, through providing more insight into the true turbulent statistics within turbulent spots. The statistics are compared to the lowest fully-turbulent DNS available in the literature to identify how the turbulent spots develop and form the fully-turbulent state. [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