Bulletin of the American Physical Society
70th Annual Meeting of the APS Division of Fluid Dynamics
Volume 62, Number 14
Sunday–Tuesday, November 19–21, 2017; Denver, Colorado
Session D24: Turbulent Boundary Layers: Wall ModelingBoundary Layers Turbulence
|
Hide Abstracts |
Chair: Alfredo Pinelli, City, University of London Room: 703 |
Sunday, November 19, 2017 2:15PM - 2:28PM |
D24.00001: The Prominent Role of the Upstream Conditions on the Large-scale Motions of a Turbulent Channel Flow Luciano Castillo, Suranga Dharmarathne, Murat Tutkun, Nicholas Hutchins In this study we investigate how upstream perturbations in a turbulent channel flow impact the downstream flow evolution, especially the large-scale motions. Direct numerical simulations were carried out at a friction Reynolds number, $Re_{\tau}=394$. Spanwise varying inlet blowing perturbations were imposed at 1$\pi h$ from the inlet. The flow field is decomposed into its constituent scales using proper orthogonal decomposition. The large-scale motions and the small-scale motions of the flow field are separated at a cut-off mode number, $M_c$. The cut-off mode number is defined as the number of the mode at which the fraction of energy recovered is $55\%$. It is found that Reynolds stresses are increased due to blowing perturbations and large-scale motions are responsible for more than $70\%$ of the increase of the streamwise component of Reynolds normal stress. Surprisingly, $90\%$ of Reynolds shear stress is due to the energy augmentation of large-scale motions. It is shown that inlet perturbations impact the downstream flow by means of the LSM. [Preview Abstract] |
Sunday, November 19, 2017 2:28PM - 2:41PM |
D24.00002: Toward models for fluctuating wall quantities in incompressible turbulent flows Aaron Towne, Xiang Yang, Parviz Moin Wall models for large-eddy simulation have been developed that provide accurate estimates of mean wall quantities such as shear stress, heat transfer, and pressure. However, these models typically do not deliver accurate predictions of the space-time fluctuations of these quantities. In this presentation, we describe some first steps toward constructing new wall models that predict the spatiotemporal properties of wall quantities by taking advantage of recent advances in our ability to identify and model the coherent structures that are known to play a central role in the near-wall dynamics. We first analyze data from a direct numerical simulation of a channel at $Re_{\tau} = 1000$ using spectral estimation techniques to isolate the contribution from different scales to fluctuating wall quantities and correlation analysis to link different spatial locations. Then, we explore how modes obtained via singular value decomposition of the resolvent operator, which is obtained from the linearized flow equations, could be used to model these fluctuations. This analysis provides a starting point for leveraging these model reduction ideas to improve the prediction of near-wall fluctuations using wall-modelled large-eddy simulation. [Preview Abstract] |
Sunday, November 19, 2017 2:41PM - 2:54PM |
D24.00003: Inner-outer predictive wall model for wall-bounded turbulence in hypersonic flow M. Pino Martin, Clara M. Helm The inner-outer predictive wall model of Mathis et al. (JFM 2011) is modified for hypersonic turbulent boundary layers. The model is based on a modulation of the energized motions in the inner layer by large scale momentum fluctuations in the logarithmic layer. Using direct numerical simulation (DNS) data of turbulent boundary layers with free stream Mach number 3 to 10, it is shown that the variation of the fluid properties in the compressible flows leads to large Reynolds number (Re) effects in the outer layer and facilitate the modulation observed in high Re incompressible flows. The modulation effect by the large scale increases with increasing free-stream Mach number. The model is extended to include spanwise and wall-normal velocity fluctuations and is generalized through Morkovin scaling. Temperature fluctuations are modeled using an appropriate Reynolds Analogy. Density fluctuations are calculated using an equation of state and a scaling with Mach number. DNS data are used to obtain the universal signal and parameters. The model is tested by using the universal signal to reproduce the flow conditions of Mach 3 and Mach 7 turbulent boundary layer DNS data and comparing turbulence statistics between the modeled flow and the DNS data. [Preview Abstract] |
Sunday, November 19, 2017 2:54PM - 3:07PM |
D24.00004: Aerodynamic heating effects on wall-modeled large-eddy simulations of high-speed flows Xiang Yang, Javier Urzay, Parviz Moin Aerospace vehicles flying at high speeds are subject to increased wall-heating rates because of strong aerodynamic heating in the near-wall region. In wall-modeled large-eddy simulations (WMLES), this near-wall region is typically not resolved by the computational grid. As a result, the effects of aerodynamic heating need to be modeled using an LES wall model. In this investigation, WMLES of transitional and fully turbulent high-speed flows are conducted to address this issue. In particular, an equilibrium wall model is employed in high-speed turbulent Couette flows subject to different combinations of thermal boundary conditions and grid sizes, and in transitional hypersonic boundary layers interacting with incident shock waves. Specifically, the WMLES of the Couette-flow configuration demonstrate that the shear-stress and heat-flux predictions made by the wall model show only a small sensitivity to the grid resolution even in the most adverse case where aerodynamic heating prevails near the wall and generates a sharp temperature peak there. In the WMLES of shock-induced transition in boundary layers, the wall model is tested against DNS and experiments, and it is shown to capture the post-transition aerodynamic heating and the overall heat transfer rate around the shock-impingement zone. This work is supported by AFOSR. [Preview Abstract] |
Sunday, November 19, 2017 3:07PM - 3:20PM |
D24.00005: The cause and resolution of log-layer mismatch in wall-modeled LES: a new perspective and its implication in complex flows George Park, Xiang Yang, Parviz Moin Log-layer mismatch (LLM) refers to the erroneous shifts of the mean velocity profile in the log-law region when wall models are coupled to the LES solution at the first off-wall grid points. It is often believed that the discretization error and subgrid-scale modeling error in the highly under resolved near-wall region contaminates the first off-wall LES solution, thereby providing inaccurate input to wall models resulting in inaccurate wall shear stress. Placing the LES/wall-model interface a couple of cells away from the wall has been recommended to avoid LLM (Kawai and Larsson, Phys. Fluids 24, 015105 (2012)). However, its non-local nature render this method impractical for flows involving complex geometry, by incurring significant overhead in LES mesh preparation and wall-model implementation. We propose an alternative remedy for LLM which warrants the removal of LLM while utilizing the first off-wall LES data. The method is based on filtering the wall-model input either in space or in time. It is simple, easy to implement, and would be particularly well suited for unstructured-grid LES involving complex geometries. We also demonstrate that LLM is caused by excessive correlation between the wall-model input and its wall shear stress output. [Preview Abstract] |
Sunday, November 19, 2017 3:20PM - 3:33PM |
D24.00006: Two-dimensional coherence spectra and their prediction using a linear Navier-Stokes based model Anagha Madhusudanan, Simon Illingworth, Ivan Marusic Recent studies of one-dimensional coherence spectra in turbulent boundary layers has helped in understanding the structure of wall-turbulence further [Baars et al., 2016, 2017]. In the present work we study these coherence spectra in further detail in two parts. First, we study the trends in the two-dimensional coherence spectra in a turbulent channel, following the analysis of Jim\'{e}nez et al. [2004] that considers the coherence of structures within the buffer layer. And secondly, this study focuses on the prediction of this coherence spectrum using a linear Navier-Stokes based model.\\ References:\\ W. J. Baars, N. Hutchins, and I. Marusic. Spectral stochastic estimation of high-reynolds-number wall-bounded turbulence for a refined inner-outer interaction model. Physical Review Fluids, 1(5):054406, 2016.\\ W. J. Baars, N. Hutchins, and I. Marusic. Self-similarity of wall-attached turbulence in boundary layers. Journal of Fluid Mechanics, 823, 2017.\\ J. Jim\'{e}nez, J. C. Del Alamo, and O. Flores. The large-scale dynamics of near-wall turbulence. Journal of Fluid Mechanics, 505:179–199, 2004.\\ [Preview Abstract] |
Sunday, November 19, 2017 3:33PM - 3:46PM |
D24.00007: Wall turbulence under the influence of Langmuir supercells in shallow water Bing-Qing Deng, Anqing Xuan, Lian Shen Langmuir circulations generated by the interaction between surface water waves and wind-driven shear turbulence are pairs of counter-rotating vortices approximately aligned along the wind direction. In shallow water, Langmuir circulations are also called Langmuir supercells, because they can engulf the whole water column and interact with the bottom boundary layer. We performed wall-resolved large-eddy simulations of Langmuir turbulence in shallow water based on the C-L equations at different Reynolds numbers ($Re$), turbulent Langmuir numbers ($La_t$), and surface wave numbers ($kh$) for a comprehensive analysis of the impact of Langmuir supercells on turbulence in the bottom boundary layer. The mean velocity and Reynolds stresses are investigated to show the effects of the non-dimensional parameters ($Re$, $La_t$, and $kh$). The budgets of Reynolds stresses are analyzed to elucidate the interaction between Langmuir supercells and other residual turbulent motions. [Preview Abstract] |
Sunday, November 19, 2017 3:46PM - 3:59PM |
D24.00008: Self-similar streak instability in the logarithmic region of turbulent channel Matteo De Giovanetti, Andrea Cassinelli, Yongyun Hwang In the present study, we report our findings on the instability of the elongated streaky motions in the logarithmic layer, and its role in the generation of vortical structures in the logarithmic layer. We perform an LES-based numerical experiment, up to $Re_\tau\approx2000$, in which a streamwise-uniform streaky motion is driven by means of an optimal forcing profile, obtained from linear theory. The spanwise spacing of the forced streaks ranges from $\lambda_z/h=0.3$ to $\lambda_z/h=1$. For a large enough amplification, turbulent statistics show characteristics consistent with a sinuous-mode instability of the streak. Furthermore, additional energy is now carried by the cross-streamwise velocity components, at a streamwise wavelength similar to the length scale of the streak instability, i.e. $\lambda_x=1\sim2\lambda_z$, where $\lambda_z$ is the spanwise spacing of the streak. Dynamic mode decomposition shows that these two structures, i.e. the streak and the vortices, both belong to the most energetic eigenstructure. The statistical description of the vortical structures resembles the characteristics of the quasi-streamwise vortices in the logarithmic layer. [Preview Abstract] |
Sunday, November 19, 2017 3:59PM - 4:12PM |
D24.00009: Investigation of Prandtl's secondary motions in a high Reynolds number duct flow Alfredo Pinelli, Alessandro Monti, Mohammad Omidyeganeh, Marco Rosti We have performed a series of wall-resolved Large Eddy Simulations of the flow in a long square duct at moderately high Reynods number (i.e., length $30~h$, $h$ being the duct semi-height and $Re_b=U_{bulk} h/\nu \simeq 20000$). Under this condition large and very large scale structures starts to appear as shown by the bimodal character of the pre-multiplied spectra of the streamwise velocity fluctuations. Apart from giving a full characterization of the large scale structures in a flow that presents two non-homogeneous directions we also investigate the pattern and origin of the secondary mean flow. In particular, following early results obtained by DNS at low Reynolds numbers that have put forward the hypothesis that the typical pattern of the secondary motion in a duct is mainly determined by the the footprints of the statistically preferred location of streamwise oriented coherent structures (Pinelli et al. JFM 644, 2010), we investigate the impact of large and very large scale coherent motions on the structure of the mean motion in a duct flow when scale separation starts to appear. A number of numerical experiments is also carried out with the aim of determining the eventual self sustenance of a flow that lacks close-to-the-wall coherent motions. [Preview Abstract] |
Sunday, November 19, 2017 4:12PM - 4:25PM |
D24.00010: Linear estimation of coherent structures in wall-bounded turbulence at Re$_\tau = 2000$ Stephan Oehler, Simon Illingworth The estimation problem for a fully-developed turbulent channel flow at Re$_\tau = 2000$ is considered. Specifically, a Kalman filter, designed using a Navier-Stokes based linear model, is given time-resolved velocity measurements at one wall-normal height (or shear stress measurements) to estimate the time-resolved velocity field at other wall-normal heights. We apply the filter to DNS simulations and compare the estimate with the true flow fields. The results show how estimator performance is affected by measurement type, measurement location, and the size of the estimated structure. Finally, we apply the filter to the linear model itself to see if the estimator performance can be predicted. [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