Bulletin of the American Physical Society
62nd Annual Meeting of the APS Division of Fluid Dynamics
Volume 54, Number 19
Sunday–Tuesday, November 22–24, 2009; Minneapolis, Minnesota
Session BA: Turbulent Boundary Layers: DNS/LES |
Hide Abstracts |
Chair: Xiaohua Wu, Royal Military College of Canada Room: 101A |
Sunday, November 22, 2009 10:30AM - 10:43AM |
BA.00001: High-Reynolds number turbulent boundary layers studied by numerical simulation Philipp Schlatter, Qiang Li, Geert Brethouwer, Arne V. Johansson, Dan S. Henningson Direct and large-eddy simulations (DNS and LES) of spatially developing high-Reynolds number turbulent boundary layers ($Re_\theta$ up to 4300) under zero pressure gradient are studied. The inflow of the computational domain and the tripping of the boundary layer is located at low Reynolds numbers $Re_ \theta \approx350$, a position where natural transition to turbulence can be expected. The simulation thus includes the spatial evolution of the boundary layer for an extended region, providing statistics and budget terms at each streamwise position. The data is obtained with up to $ O(10^{10})$ grid points using a parallelised, fully spectral method. The DNS and LES results are critically evaluated and validated, in comparison with other relevant data, \emph{e.g.}\ the experiments by \"Osterlund \emph{et al.}\ (1999). Quantities difficult or even impossible to measure, \emph{e.g.}\ pressure fluctuations and complete Reynolds stress budgets, shall be discussed. In addition, special emphasis is put on a further quantification of the large-scale structures appearing in the flow, and their relation to other wall-bounded flow as \emph{e.g.}\ channel flow. The results clearly show that with today's computer power Reynolds numbers relevant for industrial applications can be within reach for DNS/LES. [Preview Abstract] |
Sunday, November 22, 2009 10:43AM - 10:56AM |
BA.00002: Direct numerical simulation of flat-plate turbulent boundary layer up to $Re_{\theta} =2000$ Xiaohua Wu, Parviz Moin Recent flow visualization and statistical results from DNS of flat plate turbulent boundary layer are presented. We have extended the DNS of Wu {\&} Moin (JFM, \textbf{630}, 2009) to twice the streamwise length with much increased Reynolds number. The calculations were carried out using a grid of 8192$\times $500$\times $256 along the streamwise, wall-normal and spanwise directions, respectively. The computational domain in the spanwise direction was increased by 50{\%} over the earlier calculations of Wu {\&} Moin. Well-controlled boundary layer bypass transition to turbulence is promoted by the passage of isotropic turbulent flow patches in the free-stream. As in Wu {\&} Moin, in the transitional region, the instantaneous flow fields are vividly populated by hairpin vortices with distinct regularity. Many of these transitional hairpin vortices reach the local boundary layer edge. Throughout the turbulent region, forests of hairpin vortices are found to persist. As Reynolds number increases with downstream distance, the hairpin vortices exhibit more chaotic characteristics in the form of twisting and merging with more pronounced irregularity compared to their transitional counterparts. At higher Reynolds numbers hairpin forests are generally limited to about 70 percent of the local boundary layer thickness. Mean and second-order turbulence statistics and the corresponding spatial and time scales will also be presented. [Preview Abstract] |
Sunday, November 22, 2009 10:56AM - 11:09AM |
BA.00003: Effects of Computation Pipe Length on Turbulence Statistics using DNS of Turbulent Pipe Flow Cheng Chin, Andrew Ooi, Ivan Marusic, Hugh Blackburn Direct numerical simulation (DNS) of fully developed turbulent pipe flow is carried out at Re$_{\tau }\approx $ 170 and 500 (based on friction velocity, u$_{\tau }$, and pipe radii, $\delta )$ to investigate the effects of computational pipe length on the turbulence statistics. Here the DNS uses a spectral scheme in the streamwise and azimuthal directions. Various turbulence statistics are compared for different pipe lengths including the mean flow, Reynolds stresses, correlations, one-dimensional energy spectra, and skewness and flatness.The results show that in the near wall region (below the buffer region, z$^{+} \quad \le $ 30 say), a required pipe length of at least O(3000) viscous wall units is required for all turbulent statistics to converge and be independent of the length. In the outer region, comparison of spectra suggest that pipe length of 6$\pi \delta $ is sufficient for the results to be pipe-length independent. Preliminary results for higher order statistics suggest that longer lengths may be required. [Preview Abstract] |
Sunday, November 22, 2009 11:09AM - 11:22AM |
BA.00004: Fluctuating Vorticity in Turbulent Wall Layers Ronald Panton DNS data for the correlations of fluctuating vorticity in the streamwise $<\omega _{x}\omega _{x}>$, spanwise $<\omega _{z}\omega _{z}>$, and normal $<\omega _{y}\omega _{y}>$ directions is given in the various papers of Del Alamo, Jimenez, Zandonade, Moser, and Hoyas (PofF \textbf{15}, L-41; JFM, \textbf{500},p135, PofF, \textbf{18}, 011702 at four Reynolds numbers. Previously, APS Bulletin \textbf{53,} 18, 2008 EA.00004, the inner wall region was considered. It was shown that the normal component profiles at different Reynolds numbers collapse together when scaled as $<\omega _{y}\omega _{y}>$ / ($ u_{\ast}^{4}/\nu ^{2})$. However, the other components, $<\omega _{x}\omega _{x}>$ and $<\omega _{z}\omega _{z}>$, require a two-term expansion of the form \textit{F$\sim $F}$_{0}+F_{1}{\cdot}u_{\ast }/U$. The first term scaling as $<\omega \omega >_{0}$ / (($ u_{\ast }/\nu )^{2} u_{\ast}U)$ and the second scaling as $<\omega \omega >_{1}$ / ($ u_{\ast}^{4}/\nu ^{2})$. In the outer region a completely different scaling is required. An analysis of the matching behavior between the two regions shows that the common part is a function that decreases as $1/y$. This implies that in the outer region the proper scaling is $<\omega \omega >$ / [$ u_{\ast}^{3}/(h\nu )$]. Indeed, profiles of all three components collapse in the outer region in this variable. Furthermore, all three components show a marked tendency toward the same level and isotropic behavior. [Preview Abstract] |
Sunday, November 22, 2009 11:22AM - 11:35AM |
BA.00005: DNS of Turbulent Boundary Layers with/without External Pressure Gradient Based on a Multi-Scale, Dynamic Recycle Inflow Condition Approach Luciano Castillo, Juan G. Araya, Charles Meneveau, Kenneth Jansen A method for prescribing realistic turbulent velocity inlet boundary conditions is presented for simulations of spatially evolving turbulent boundary layers. The standard rescaling process requires prior knowledge about how the appropriate velocity and length scales are related between the inlet and recycle stations. In the present study the scales for the inner and outer regions are determined from the multi-scale approach based on the original equilibrium similarity method developed by Castillo and George (2001) (for PG flows). In addition, a new dynamic approach is proposed in which power law ratios of inner/outer scales are used with scaling exponents that may depend on flow conditions and are deduced dynamically by involving an additional plane, a ``test plane''. This improvement, as well as the use of multiple velocity scales, permits the simulations of turbulent boundary layers subjected to arbitrary pressure gradients. DNS for zero (ZPG), and pressure gradient flows (APG and FPG) are discussed with special emphasis on adverse pressure gradient flows. In addition, new simulations at about $R_{\theta} \simeq$ 3,000 will be shown and compared with experimental data. [Preview Abstract] |
Sunday, November 22, 2009 11:35AM - 11:48AM |
BA.00006: Effects of Mach Number on Near-Wall Turbulence Structures in Supersonic Turbulent Boundary Layers Maher Lagha, John Kim, Jeff Eldredge, Xiaolin Zhong In order to examine the effects of high Mach number on turbulence structures in compressible turbulent boundary layers, we performed direct numerical simulation of a spatially evolving supersonic turbulent boundary layer. A hybrid numerical method coupling the 5th-order WENO scheme with a 5th-order upwind finite-difference scheme was used in order to compute turbulence accurately while capturing sharp gradients that might exist in high Mach number flows. The rescale-and-recycle method was used for the inflow boundary condition. The Mach numbers were varied from 2.5 to 7 while the Reynolds numbers based on the boundary-layer displacement, wall-shear velocity and the viscosity at the wall were kept about the same. In order to determine the relative importance of the true compressibility effect and the variable-property effect in high Mach number boundary layers, we also solved modified Navier-Stokes equations, in which the energy equation was modified to include a heat sink term (Coleman, Kim and Moser, J. Fluid Mech. 305). The effects of Mach number on turbulence statistics as well as an analysis of the near-wall turbulence structures will be presented. [Preview Abstract] |
Sunday, November 22, 2009 11:48AM - 12:01PM |
BA.00007: Direct numerical simulations of hypersonic boundary layers varying wall-to-freestream temperature ratio Lian Duan, Izaak Beekman, Pino Martin The effects of wall-temperature condition on the statistics of compressible turbulent boundary layers are investigated using direct numerical simulation (DNS). DNS of turbulent boundary layers at Mach 5 with the ratio of wall-to-edge temperature Tw/Te from 1.0 to 5.4 (Cases M5T1 through M5T5) are performed. Case M5T5 corresponds to nearly adiabatic wall, and cases M5T1 through M5T4 correspond to cooled isothermal walls. The validity of Morkovin's scaling, Walz's equation, and the standard and modified strong Reynolds analogy are assessed. Turbulent kinetic energy, contours of spanwise vorticity, near-wall streaks, and two-point correlations show that the temperature cooling stabilizes the turbulence in compressible boundary layers. Compressibility effects are enhanced by wall cooling but remain secondary to the dynamics already appearing in incompressible flow, and the turbulence dissipation remains primarily solenoidal. Hairpin packets are stronger and more coherent for colder wall simulations, while the average hairpin angle remains insensitive to wall temperature. [Preview Abstract] |
Sunday, November 22, 2009 12:01PM - 12:14PM |
BA.00008: Large eddy simulation of boundary layers with embedded spanwise vortices Iftekhar Naqavi, Ugo Piomelli We performed large eddy simulations of a zero-pressure-gradient boundary layer interacting with strong, spanwise-oriented vortices. A periodic array of spanwise vortices is generated at the edge of the boundary layer; as they advect downstream, they introduce strong perturbations that extend to the near-wall region, and alter significantly the turbulence dynamics. Localized separation regions are observed below the vortices. Phase averaged data show that the vortices have lost their coherence $10\delta$ downstream of the point where they are generated however the boundary layer does not recover its equilibrium state even after $60\delta$. While the vortices remain coherent the phase-averaged velocity profiles show a strong wake, and the logarithmic law is shifted first upwards, and then downwards as the vortex passes. After the coherence is lost all the phases show almost identical behaviour, but the recovery towards an equilibrium boundary layer profile is slow. The fluctuations due to the vortices can be decomposed into a periodic and a random component. The periodic fluctuations supply up to 50\% \ of the total Reynolds stresses in the coherent region, but their contribution is reduced significantly when the vortices decay, and the Reynolds stress due to the random fluctuations is dominant. Flow visualizations show strong uplift and stretching of the near-wall vortices. [Preview Abstract] |
Sunday, November 22, 2009 12:14PM - 12:27PM |
BA.00009: Large-eddy simulation investigation of large-scale structures in a long channel flow Daniel Chung, Beverley McKeon We report statistics of large-scale near-wall structures from large-eddy simulation (LES) of turbulent channel flow at friction Reynolds numbers 2 k and 200 k. To properly assess the behavior of large-scale structures, we perform simulations in a channel whose length is 96 times its half-channel width. In agreement with experiments, these large-scale structures were found to modulate the statistics of the underlying small-scale fluctuations. In particular we report that, near the wall, large-scale high-speed streaks carry more intense superimposed small-scale fluctuations, but that this correlation is reversed away from the wall. We also report that the convection velocity of these large-scales near the wall departs slightly, but unequivocally, from the mean velocity. [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