Bulletin of the American Physical Society
69th Annual Meeting of the APS Division of Fluid Dynamics
Volume 61, Number 20
Sunday–Tuesday, November 20–22, 2016; Portland, Oregon
Session G34: Turbulence: Modeling via Parameterization and Analysis |
Hide Abstracts |
Chair: Robert Moser, University of Texas at Austin Room: Oregon Ballroom 203 |
Monday, November 21, 2016 8:00AM - 8:13AM |
G34.00001: Hybrid LES/RANS simulation of a turbulent boundary layer over a rectangular cavity Qi Zhang, Sigfried Haering, Todd Oliver, Robert Moser We report numerical investigations of a turbulent boundary layer over a rectangular cavity using a new hybrid RANS/LES model [1] and the traditional Detached Eddy Simulation (DES). Our new hybrid method aims to address many of the shortcomings from the traditional DES. In the new method, RANS/LES blending controlled by a parameter that measures the ratio of the modeled subgrid kinetic energy to an estimate of the subgrid energy based on the resolved scales. The result is a hybrid method automatically resolves as much turbulence as can be supported by the grid and transitions appropriately from RANS to LES without the need for ad hoc delaying functions that are often required for DES. Further, the new model is designed to improve upon DES by accounting for the effects of grid anisotropy and inhomogeneity in the LES region. We present comparisons of the flow features inside the cavity and the pressure time history and spectra as computed using the new hybrid model and DES. [Preview Abstract] |
Monday, November 21, 2016 8:13AM - 8:26AM |
G34.00002: The Use of the Reynolds Stress Transport Equation to Constrain Eigenvectors Perturbations on Model Form UQ of RANS Simulations Luiz Sampaio, Roney Thompson, Wouter Edeling, Aashwin Mishra, Gianluca Iaccarino Despite the recent developments in LES and DNS approaches for turbulent flow simulations, RANS modeling is still vastly used by industry, due to its inherent low cost. Since accuracy is a concern in RANS modeling, model-form UQ is an essential tool for assessing the impacts of this uncertainty on quantities of interest. Bounding values for the eigenvalues of the dimensionless deviatoric part of the Reynolds Stress tensor (RST) can be obtained from realizability constraints, and therefore can be used as a first step towards a general perturbation approach. In this connection, decoupling the perturbation into an intensity (kinetic energy), a shape (eigenvalues), and an orientation (eigenvectors) parts constitutes a natural methodology to evaluate the model form UQ associated to the RST modeling. In this work, we show that ignoring eigenvectors perturbations can lead to significant impacts on the results from the UQ analysis. Besides that, we use the RST Equation as a constraint to impose some consistency between eigenvectors and eigenvalues perturbations, where the latter can be obtained from a more standard technique. We applied this methodology on the convex channel flow, and show the benefits of including the eigenvectors perturbations predicted by this methodology. [Preview Abstract] |
Monday, November 21, 2016 8:26AM - 8:39AM |
G34.00003: A dynamic hybrid subgrid-scale modeling framework for large eddy simulations Romit Maulik, Omer San We put forth a dynamic modeling framework for sub-grid parameterization of large eddy simulation of turbulent flows based upon the use of the approximate deconvolution (AD) procedure to compute the eddy viscosity constant self-adaptively from the resolved flow quantities. In our proposed framework, the test filtering process of the standard dynamic model is replaced by the AD procedure and a posteriori error analysis is performed. The robustness of the model has been tested considering the Burgers, Kraichnan, Kolmogorov turbulence problems. Our numerical assessments for solving these canonical decaying turbulence problems show that the proposed approach could be used as a viable tool to address the turbulence closure problem due to its flexibility. [Preview Abstract] |
Monday, November 21, 2016 8:39AM - 8:52AM |
G34.00004: Axisymmetric Afterbody Test Case for CFD Validation Kevin Disotell, Christopher Rumsey As simulation complexity increases, the corresponding need for systematic, high-fidelity validation data sets continues to be important to advance physics-based CFD models. To this end, a parametric body of revolution is proposed as an experimental platform to support a wide validation domain for turbulent boundary layers outside the current bounds of DNS. Recognizing the challenges of detailed flow exploration on complex 3-D geometries, an analytically-defined body of revolution is pursued as a tractable, state-of-the-art measurement case for complex turbulent flows having extra rates of strain. The central feature of the concept based upon work by Presz Jr. \& Pitkin [J. Aircraft \textbf{11}, 677 (1974)] is an interchangeable afterbody which can be tailored to distort a turbulent boundary layer in various ways, with incoming properties controlled by the forebody. An introduction to the test case design and overview of recent progress focused on smooth-body, turbulent separation physics are presented. [Preview Abstract] |
Monday, November 21, 2016 8:52AM - 9:05AM |
G34.00005: Nonlinear forcing in the resolvent analysis of wall-turbulence Kevin Rosenberg, Adrian Lozano Duran, Aaron Towne, Beverley McKeon The resolvent analysis of McKeon and Sharma (JFM, 2010) formulates the Navier-Stokes equations as an input/output system in which the nonlinearity is treated as a forcing that acts upon the linear dynamics to yield a velocity response across wavenumber/frequency space. DNS data for a low Reynolds number turbulent channel ($Re_{\tau} = 180 $ ) is used to investigate the structure of the nonlinear forcing directly. Specifically, we explore the spatio-temporal scales where the forcing is active and analyze its interplay with the linear amplification mechanisms present in the resolvent operator. This work could provide insight into self-sustaining processes in wall-turbulence and inform the modeling of scale interactions in large eddy simulations. [Preview Abstract] |
Monday, November 21, 2016 9:05AM - 9:18AM |
G34.00006: An anisotropic subgrid stress model for high aspect ratio grids Robert Moser, sigfried haering Standard algebraic eddy viscosity subgrid stress models are formulated based on scalar measures of the local grid, and implicitly assume that the resolution is isotropic. However, complex simulation domains and computational costs associated with problems of engineering interest often necessitate grids with high aspect ratio cells. We present an anisotropic extension of Metias and Lesieur's structure function subgrid stress model [1]. Unlike existing algebraic SGS models, this model is constructed directly through the composition of resolution and resolved turbulence anisotropy. Comparisons with filtered DNS of forced isotropic homogeneous turbulence show the model to significantly outperform general isotropic SGS models with increasing resolution anisotropy. \newline \newline [1] O. Metais and M. Lesieur, ``Spectral large-eddy simulation of isotropic and stably stratified turbulence,'' Journal of Fluid Mechanics, vol. 239, 1992 [Preview Abstract] |
Monday, November 21, 2016 9:18AM - 9:31AM |
G34.00007: Scaling analysis on filtered near wall turbulence Prakash Mohan, Robert Moser Large Eddy Simulations (LES) directly represent larger scale turbulent motions and model the effects of small scale motions. However in the near wall region the large dynamically important eddies scale in viscous wall units, which makes resolving them in a high Reynolds number LES very expensive. This motivates the use of wall-modeled LES, in which these near-wall eddies are modeled. To aid in the development of new wall models, we pursue an asymptotic analysis of the filtered Navier-Stokes equations, in the limit in which the horizontal filter scale is large compared to the thickness of the wall layer. It will be shown that in this limit the filtered velocities $\bar{u}$ and subgrid stresses $\bar{\tau}$ in the near-wall layer are determined to zeroth order by filtered velocities at the boundary of the wall layer. Further the asymptotics suggest that there is a scaled universal velocity profile $f$ and subgrid stress profile $g$ in the near-wall region. The validity of this result will be tested and the profiles $f$ and $g$ will be evaluated through analysis of DNS data from channel flow at $Re_{\tau}=5200$. [Preview Abstract] |
Monday, November 21, 2016 9:31AM - 9:44AM |
G34.00008: Modeling and simulation of turbulence not at statistical equilibrium Arash Nouri Gheimassi, Peyman Givi, William Layton Standard eddy viscosity models, while popular, cannot represent backscatter and have severe difficulties with complex turbulence not at statistical equilibrium.~In this work, we give a derivation of eddy viscosity models from an equation for the evolution of variance in a turbulent flow showing how to correct eddy viscosity models. We show the corrected models preserve important features of the true Reynolds stresses and give algorithms for their discretization including a minimally invasive modular step to adapt an eddy viscosity computational~methodology to the extended models. A numerical test is given with the usual and over diffusive Smagorinsky model. The consistency and the predictive capability of the model are established by conducting LES of a three-dimensional compressible mixing layer, and comparison with DNS data. The correction does~exhibit intermittent backscatter. [Preview Abstract] |
Monday, November 21, 2016 9:44AM - 9:57AM |
G34.00009: Analytical study on the SGS force around an elliptic Burgers vortex Hiromichi Kobayashi The subgrid-scale (SGS) force around an elliptic Burgers vortex is analytically examined. In turbulence, there are a lot of vortex-tubes whose cross sections are known to be approximated as the ellipse. In this study, the biaxial elliptic Burgers vortex is produced by adding the compressive and extensional background straining flow to the conventional Burgers vortex. By using a filtering operation, we revealed that the energy transfer by the Reynolds stress term applying the Bardina model exhibits negative correlation to that by the true SGS stress term. However, it has been recently reported that a combination of the Bardina Reynolds term and the eddy viscosity model gives good performance even for the coarse LES of turbulent channel flows. In order to understand that, we discuss some SGS forces: by the true SGS stress tensor, by the eddy viscosity model, by the modified Leonard term and by the Bardina Reynolds term. [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