Bulletin of the American Physical Society
68th Annual Meeting of the APS Division of Fluid Dynamics
Volume 60, Number 21
Sunday–Tuesday, November 22–24, 2015; Boston, Massachusetts
Session L22: Turbulent Boundary Layers IV |
Hide Abstracts |
Chair: Julian Hunt, University College London Room: 210 |
Monday, November 23, 2015 4:05PM - 4:18PM |
L22.00001: Thin shear layers in homogeneous high Reynolds number turbulence and in turbulent boundary layers Takashi Ishihara, Koji Morishita, Julian Hunt Direct numerical simulations (DNSs) at high Reynolds number show for forced homogeneous isotropic turbulence at $R_\lambda\sim 1000$ that randomly moving, strong thin shear- layers form in the interior (T/In), in which there are high-enstrophy, micro-scale vortex tube structures. These layers have thicknesses of the order of the Taylor micro-scale and the interfaces at the outside of the layers act as a partial barrier to the fluctuations on either side of the layers. In the turbulent boundary layers (TBL) at $R_\lambda\sim 100$, conditional statistics show three different types of thin shear layers; at the outer edge (T/NT), in the interior (T/In) and within the buffer layer near the wall (T/W). These layers act as barriers to the fluctuations on either side and can have controlling effects on the overall flow. The internal and external characteristics and role of the thin shear layers in homogeneous turbulence and in TBL are compared. [Preview Abstract] |
Monday, November 23, 2015 4:18PM - 4:31PM |
L22.00002: General mechanisms of thin layers in high Reynolds number turbulent flows Julian Hunt, Takashi Ishihara, Koji Morishita Mechanisms and computation are presented for the three types of thin, high vorticity, randomly moving shear layers at high Reynolds number. They decorrelate eddy motions on each side and, in the first two types, have an internal micro-scale, dissipative structure. Their form also depends on the mean strain/shear outside the layer, and the proximity of any resistive boundaries. The first type (T/NT) lie between regions of sheared turbulence and external non-turbulent motions. Depending on whether the inflection points of the conditional mean shear profile, $\langle U(y-y_i)\rangle$, relative to the interface coordinate $y_i$, are on the outside or inside edges of the layer, the forms of the interface are ``nibbling'' motions on the scale of the layer thickness or large ``engulfing'' motions, which affect the overall flow structure. In the second type (T/In), which occurs in the interior of turbulent flows, because the interface instabilities are suppressed, the stretching increases more than in T/NT, causing the micro-scale vorticity, velocity and dissipation to greatly exceed Kolmogorov's theory. The third type (T/W) within the buffer wall layer, by blocking outer eddies, determines the displaced form of the mean logarithmic profile, and fluctuations of wall shear stress. [Preview Abstract] |
Monday, November 23, 2015 4:31PM - 4:44PM |
L22.00003: Passive scalars in turbulent channel flow at high Reynolds number Sergio Pirozzoli, Matteo Bernardini, Paolo Orlandi We study passive scalars in turbulent plane channels at computationally high Reynolds number, which allows to observe previously unnoticed effects. The mean scalar profiles are found to obey a generalized logarithmic law which includes a linear correction term in the whole lower half-channel, and they follow a universal parabolic defect profile in the core region. This is consistent with recent findings regarding the mean velocity profiles in channel flow. The scalar variances also exhibit a near universal parabolic distribution in the core flow, and hints of a sizeable log layer, unlike the velocity variances. The energy spectra highlight the formation of large scalar-bearing eddies spanning each half-channel, which are caused by production excess over dissipation, and which are clearly visible in the flow visualizations. Close correspondence of the velocity and scalar eddies is observed, the main difference being that the latter have more convoluted interfaces, which translates into higher scalar dissipation. Another notable Reynolds number effect is the decreased correlation of the scalar field with the vertical velocity field, which is traced to the reduced effectiveness of ejection events. [Preview Abstract] |
Monday, November 23, 2015 4:44PM - 4:57PM |
L22.00004: Pressure gradient influence in turbulent boundary layers Nico Reuther, Christian J. Kaehler Understanding wall-bounded turbulence is still an ongoing process. Although remarkable progress has been made in the last decades, many challenges still remain. Mean flow statistics are well understood in case of zero pressure gradient flows. However, almost all turbulent boundary layers in technical applications, such as aircrafts, are subjected to a streamwise pressure gradient. When subjecting turbulent boundary layers to adverse pressure gradients, significant changes in the statistical behavior of the near-wall flow have been observed in experimental studies conducted however the details dynamics and characteristics of these flows has not been fully resolved. The sensitivity to Reynolds number and the dependency on several parameters, including the dependence on the pressure gradient parameter, is still under debate and very little information exists about statistically averaged quantities such as the mean velocity profile or Reynolds stresses. In order to improve the understanding of wall-bounded turbulence, this work experimentally investigates turbulent boundary layer subjected to favorable and adverse pressure gradients by means of Particle Image Velocimetry over a wide range of Reynolds numbers, 4200 \textless Re$_{\mathrm{\tau }}$ \textless 13400. The contribution of the coherent structures to the mean flow statistics was found to increase significantly for a flow subjected to an adverse pressure gradient. [Preview Abstract] |
Monday, November 23, 2015 4:57PM - 5:10PM |
L22.00005: Space filling attributes of the turbulent motions responsible for the generation of the Reynolds stress Caleb Morrill-Winter, Joseph Klewicki, Jimmy Philip, Ivan Marusic The self-similar inertial layer in wall-bounded turbulence is most commonly associated with the mean velocity profile exhibiting logarithmic behavior. Recent evidence indicates that the value the leading order coefficient describing this logarithmic region (von Karman constant) is intimately related to the space-filling properties of the turbulent motions responsible for the generation of the Reynolds stress. In the present experiments, a compact hot-wire probe employed a temporally optimized processing scheme to obtain well-resolved $uv$ time series over the range $2,000 \leq \delta^+ = \delta_{99} u_\tau/\nu \leq 12,700$. Over the entire Reynolds number range good spatial and temporal resolution was maintained by utilizing the low speed, large scale attributes of the MWT and the FPF, at the University of Melbourne and University of New Hampshire, respectively. The present study examines statistical properties associated with the fraction of time, and associated length scale, that the $uv$ time series is negative. These statistics have connection to the area coverage of the motions responsible for the wall-ward transport of momentum. Results are described and interpreted in a context consistent with the structure of the mean momentum equation. [Preview Abstract] |
Monday, November 23, 2015 5:10PM - 5:23PM |
L22.00006: Wall scaling laws for turbulent BL before and after the Reynolds shear stress maxima Noor Afzal, Abu seena, Bushra Afzal The turbulent shear flow has a critical point due to maximas of Reynolds stress tensor has significant role. Three layers with length scales (inner $\nu/u_\tau$, meso $\sqrt{\nu \delta/u_\tau}$, outer $\delta$) have been analyzed. Below this crucial point the mesolayer inner limit matches with with outer limit of wall layer. Above this crucial point the outer limit of mesolayer matches with inner limit of outer layer. The log law velocity and Reynolds in two overlap regions, above and below the critical point, have been presented. The Reynolds shear stress maxima $\tau_{max}/\tau_w$ occurs at a point where ratio of mesolayer to outer lengths is of order $R_\tau^{-1/2}$ (= $\sqrt{\nu/\delta u_\tau}$), and at this point DNS and experimental data predict $U_m/U_e = 2/3$ (where Um = mesolayer velocity and Ue = velocity at boundary edge). The turbulent burst time period also scale with mesolayer time. The shape factor in a TBL shows linear behavior with non-dimensional mesolayer length scale. In special case $U_m/U_e =1/2$, is due to Izakson and Millikan. The above predictions are supported by experimental and DNS data. [Preview Abstract] |
Monday, November 23, 2015 5:23PM - 5:36PM |
L22.00007: Arrangement of scale-interaction and large-scale modulation in high Reynolds number turbulent boundary layers Woutijn J. Baars, Nicholas Hutchins, Ivan Marusic Interactions between small- and large-scale motions are inherent in the near-wall dynamics of wall-bounded flows. We here examine the scale-interaction embedded within the streamwise velocity component. Data were acquired using hot-wire anemometry in ZPG turbulent boundary layers, for Reynolds numbers ranging from $Re_{\tau} \equiv \delta U_{\tau}/\nu \approx 2800$ to 22800. After first decomposing velocity signals into contributions from small- and large-scales, we then represent the time-varying small-scale energy with time series of its instantaneous amplitude and instantaneous frequency, via a wavelet-based method. Features of the scale-interaction are inferred from isocorrelation maps, formed by correlating the large-scale velocity with its concurrent small-scale amplitude and frequency. Below the onset of the log-region, the physics constitutes aspects of amplitude modulation and frequency modulation. Time shifts, associated with the correlation extrema---representing the lead/lag of the small-scale signatures relative to the large-scales---are shown to be governed by inner-scaling. Wall-normal trends of time shifts are explained by considering the arrangement of scales in the log- and intermittent-regions, and how they relate to stochastic top-down and bottom-up processes. [Preview Abstract] |
Monday, November 23, 2015 5:36PM - 5:49PM |
L22.00008: Assessment of fluctuating pressure gradient using acceleration spectra in near wall flows Daniel Cadel, K. Todd Lowe Separation of contributions to the fluctuating acceleration from pressure gradient fluctuations and viscous shear fluctuations in the frequency domain is examined in a turbulent boundary layer. Past work leveraging turbulent accelerations for pressure gradient measurements has neglected the viscous shear term from the momentum equation---an invalid assumption in the case of near wall flows. The present study seeks to account for the influence of the viscous shear term and spectrally reject its contribution, which is thought to be concentrated at higher frequencies. Spectra of velocity and acceleration fluctuations in a flat plate, zero pressure gradient turbulent boundary layer at a momentum thickness Reynolds number of 7500 are measured using a spatially resolving three-component laser Doppler velocimeter. This canonical case data is applied for validation of the spectral approach for future application in more complex aerodynamic flows. [Preview Abstract] |
Monday, November 23, 2015 5:49PM - 6:02PM |
L22.00009: DNS of turbulence around a wing section at moderate Reynolds number Philipp Schlatter, Seyed M. Hosseini, Ricardo Vinuesa, Ardeshir Hanifi, Dan S. Henningson We present the results of a large-scale simulation of the turbulent flow around a NACA-4412 wing section. The achieved Reynolds number is $Re_c=400000$ based on the chord length ($Re_\theta=3000$ based on momentum thickness), at angle of attack of 5 degrees. The fully resolved direct numerical simulation is performed using the spectral-element code Nek5000 with 3.2 billion grid points. After discussing details of the setup, e.g. boundary conditions and flow tripping at the leading edge, the focus is on the turbulent boundary layers under favorable and adverse pressure gradient developing along the wing surfaces. A first question to address is the definition of boundary-layer thickness in curved geometries. The adverse pressure gradients (APG) remain fairly constant $\beta<4$ for the most part of the wing's upper side, only towards the trailing edge, incipient separation and much higher $\beta$ are observed. The mean profiles show typical characteristics of APG boundary layers, to which we will compare in detail. A distinct outer peak in the fluctuations can be seen. These observations will be complemented with spectral views of the growing outer-layer influence. Furthermore, visualizations of the vortical structures will be shown, both on the wing, but also in the wake region. [Preview Abstract] |
Monday, November 23, 2015 6:02PM - 6:15PM |
L22.00010: DNS of self-similar adverse pressure gradient turbulent boundary layer at incipient separation Julio Soria, Vassili Kitsios, Callum Atkinson, Juan Sillero, Guillem Borrell, Ayse Gungar, Javier Jimenez A direct numerical simulation of a self-similar adverse pressure gradient turbulent boundary layer (APG-TBL) flow at incipient separation has been carried out. The maximum Reynolds number based on the momentum thickness, $Re_{\delta_2}$, reached in this DNS is 6,500. A wall-normal far-field boundary condition to effect the desired APG that will lead to the desired self-similar flow at the verge of separation has been developed. The self-similar analysis of the mean turbulent boundary layer equations yields the necessary conditions for a self-similar mean flow to exists. These conditions are tested using the DNS APG-TBL data base. First and second order statistics of the velocity across the APG-TBL are also presented in the light of the self-similar analysis results and compared to the results of a zero pressure gradient turbulent boundary layer DNS with similar mean inflow characteristics as the APG-TBL. [Preview Abstract] |
Monday, November 23, 2015 6:15PM - 6:28PM |
L22.00011: Structural features of the $k_{x}^{-1}$ region of turbulent pipe flow at Re$_{\tau }=$3008 Junsun Ahn, Hyung Jin Sung Structural features of a turbulent pipe flow were explored by using the direct numerical simulation data at Re$_{\tau }=$3008 (Ahn et al. 2015). Based on the pre-multiplied streamwise energy spectra of the streamwise velocity fluctuations, three spectral regions were classified: the inner site, the outer site and $k_{x}^{-1}$ region. The inner site was created by the self-sustaining near-wall cycle with $\lambda_{x}^{+}\approx $1000, where $\lambda_{x}$ is the streamwise wavelength. The outer site was made due to very-large-scale motions with $\lambda_{x}$/R$\approx $10, which were generated by the streamwise pseudo-alignment of the adjacent large-scale motions. Between the inner and outer sites, the $k_{x}^{-1}$ region appeared at $y^{+}=$90-300, where $\lambda_{x}\ge $20$y$ and $\lambda_{x}$/R$\le $5. By using the conditional averaging, self-similar structures of the streamwise velocity fluctuations structures in the $k_{x}^{-1}$ region were retained, which were considered as the attached eddies proposed by Townsend (1976). In addition, the vortical structures in the $k_{x}^{-1}$ region were examined by two-point correlation of the velocity components and the vortices in order to find the dominant behavior of the structures. [Preview Abstract] |
Monday, November 23, 2015 6:28PM - 6:41PM |
L22.00012: Vortex packet recovery in a turbulent boundary layer perturbed by an array of cylinders Yan Ming Tan, Ellen Longmire PIV measurements were acquired in a zero pressure gradient turbulent boundary layer (Re$_{\tau} = $ 2500) perturbed by a narrowly spaced (0.2$\delta )$ array of cylinders. Two array heights were considered with one extending to the top of the log region and the other to the top of the boundary layer. Wall-parallel measurements were obtained at three locations in the log region by fixed and flying PIV. The measurement system for flying PIV moves with the flow to track the evolution of structures upstream and downstream of the array. Initially, both arrays disrupt the packets such that none are apparent. Then, packets appear either to recover or re-initiate at some distance downstream. A packet signature was denoted by a low momentum region bounded by counter rotating swirling structures. A low momentum region identification algorithm was applied to both fixed and flying PIV data to quantify packet recovery downstream of the array. The results indicate that packets reappear sooner further from the wall and later closer to the wall for the shorter array supporting the top down notion of packet reorganization proposed by Zheng {\&} Longmire (JFM, 2014). The opposite trend was observed for the taller array whereby packets recovered earlier closer to the wall and later further from the wall. [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. |
© 2019 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