Bulletin of the American Physical Society
71st Annual Meeting of the APS Division of Fluid Dynamics
Volume 63, Number 13
Sunday–Tuesday, November 18–20, 2018; Atlanta, Georgia
Session F21: Boundary Layers: Modeling and Analysis |
Hide Abstracts |
Chair: George Park, University of Pennsylvania Room: Georgia World Congress Center B309 |
Monday, November 19, 2018 8:00AM - 8:13AM |
F21.00001: Resolvent and Spectral POD Analyses of a Laminar Separation Bubble Hao Zhang, Alberto Padovan, Clarence Rowley, Wen Wu, Charles Meneveau, Rajat Mittal Resolvent analysis is an increasingly popular technique capable of providing insight about the input-output characteristics of a given fluid flow at a specified frequency and wavenumber. Moreover, the resolvent of the linearized Navier-Stokes operator can be used to compute the pseudospectrum, which can be important for non-normal operators as often arise in shear flows.
Direct numerical simulations are performed for three-dimensional laminar boundary layer flow in which suction and blowing are imposed at the top of the computational domain to generate a separation bubble at the bottom surface. The linear input-output (resolvent) operator is computed relative to the spanwise- and time-averaged mean flow, and the optimal forcing and response modes are obtained via a randomized singular value decomposition. The resulting forcing modes illustrate regions where the flow is most sensitive to perturbations. The corresponding response modes are compared with spectral (frequency-domain) POD modes obtained from snapshots of the direct numerical simulations, and the two sets of modes show similar features. |
Monday, November 19, 2018 8:13AM - 8:26AM |
F21.00002: Displacement thickness based recycling inflow generation method for spatially developing turbulent boundary layer simulations Samvit Kumar, Rajat Mittal, Charles Vivant Meneveau An improved method for generation of turbulent inflow for simulations of developing boundary layers is presented, based on prior recycling methods for flow over smooth (Lund et al., 1998) and rough (Yang and Meneveau, 2015) surfaces. Both these methods rely on obtaining δ99 from the mean velocity profiles based on a velocity threshold. Being dependent on the profile shape, δ99 can be noisy and can suffer from large undesirable fluctuations, despite being time averaged. To cushion the effects of unusual mean velocity profiles, a profile-integrated quantity, like the displacement thickness (δ*), can be used instead of δ99. In the recycling method, velocities on a sample plane are rescaled and recycled back to the inlet, as the inflow velocity. A surface geometry dependent, roughness-length related scale is chosen to rescale the inner layer and δ* is chosen instead of δ99 as the length scale to rescale the outer layer. The blending function, dependent on both the inner and the outer length scales, is used to combine the two profiles, to obtain the inflow velocity. Since δ* depends on the profile shape, an iterative scheme is implemented. Several applications and test cases are presented. |
Monday, November 19, 2018 8:26AM - 8:39AM |
F21.00003: Abstract Withdrawn
|
Monday, November 19, 2018 8:39AM - 8:52AM |
F21.00004: Well-posedness of the slip wall boundary conditions for the LES equations Kevin Patrick Griffin, Sanjeeb T Bose, Parviz Moin A Robin boundary condition admitting instantaneous and local transpiration (but no net transpiration) has been shown to be a suitable boundary condition for wall-modeled large eddy simulation. Well-posedness of these boundary conditions is analyzed through the consideration of the global kinetic energy and entropy evolution equations for the incompressible and compressible LES governing equations, respectively. Most existing analysis of well-posedness of wall boundary conditions consider Dirichlet boundary conditions for the wall normal velocity (no penetration) for the filtered or unfiltered Navier-Stokes equations. The present analysis seeks to identify under which conditions a bounded solution can be expected, which is particularly valuable when the slip lengths are dynamically computed. Preliminary computational results from the LES of a plane channel flow are presented to support these theoretical claims. The global kinetic energy in the domain is shown to grow or decay in time, depending on the procedure for enforcing the slip wall boundary condition. |
Monday, November 19, 2018 8:52AM - 9:05AM |
F21.00005: Predictive LES wall modeling via physics-informed neural networks Xiang Yang, Suhaib Zafar, Jianxun Wang, Heng Xiao While data-based approaches were found to be useful for sub-grid scale (SGS) modeling in Reynolds-averaged Navier-Stokes (RANS) simulations, there have not been many attempts of using machine learning (ML) techniques for wall modeling in large-eddy simulations (LES). LES wall modeling poses additional challenges to data-based modeling approaches. First, datasets of higher fidelity are not easily accessible. Second, wall modeling needs to account for both near-wall small scales and large scales above the wall. In this work, we discuss how the above-noted challenges may be addressed. We will also show the necessity of incorporating physics insights in model inputs, i.e. using inputs that are inspired by the vertically integrated thin boundary layer equations and the eddy population density scalings. We will show that the inclusion of above physics-based considerations would enhance extrapolation capabilities of a neural network to flow conditions that are not within the train data. Being cheap-to-evaluate and using only channel flow data at Re_\tau=1,000, the trained networks are found to capture the law of the wall at arbitrary Reynolds numbers and outperform the conventional equilibrium model in a non-equilibrium flow. |
Monday, November 19, 2018 9:05AM - 9:18AM |
F21.00006: Wall modeled large eddy simulation of a spatially developing non-equilibrium turbulent boundary layer Minjeong Cho, George Ilhwan Park, Parviz Moin Present study aims to examine the performance of wall models in a three-dimensional turbulent boundary layer (3DTBL). A wall modeled LES is conducted for a square-duct flow where a spatially-developing 3DTBL is generated by a 30-degree bend, following the experiment of Schwarz & Bradshaw (JFM, 1994). The cross-stream pressure gradient imposed by the bend creates a skewed velocity profile, where the surface streamlines are deflected by up to 22 degrees relative to the centerline velocity. The Reynolds number based on the momentum thickness and the free-stream velocity ranges from 4000 to 9000. An equilibrium wall model is used with a coarse mesh deploying 10 to 20 cells within the local boundary layer thickness. Prediction of mean velocities, wall pressure, and turbulence quantities from the wall modeled LES are in reasonable agreement with the experiment. Effectiveness of the equilibrium wall model in capturing non-equilibrium effects (e.g., evolution of Reynolds-stress budget in the outer layer) will be discussed. |
Monday, November 19, 2018 9:18AM - 9:31AM |
F21.00007: Effect of Favorable Pressure Gradient and Permeable Wall on the Flow Regime in the Boundary Layer Alexey Sakhnov We consider air flow in a plane convergent channel, where the acceleration parameter K remains constant over the entire channel length. The lower permeable wall has a constant wall-normal air flux jw=Vw/Ue, where negative and positive values correspond to gas suction and air injection, respectively. Such flow can be well approximated by the parabolized two-dimensional momentum and continuity equations for a steady incompressible boundary layer. These equations are supplemented with k-ω-γ turbulence model. The model is able to predict laminar-turbulent transition and flow relaminarization using intermittency factor. |
Monday, November 19, 2018 9:31AM - 9:44AM |
F21.00008: Stream-wise Periodic Direct Numerical Simulations of an Incompressible, Flat-Plate Boundary Layer: Validations and Analysis Joseph Y Ruan, Guillaume Blanquart Direct numerical simulations of an incompressible, flat-plate, turbulent boundary layer were performed via a novel simulation method, inspired by Spalart (J. Fluid Mech., vol. 187, 1988, p. 61-98) and his further simplifications on inflow recycling (Int. J. Heat Fluid Fl., vol. 27, 2006, p. 902-910). The new method is stream-wise periodic and uses normalized wall-normal coordinates and overcomes key limitations of Spalart’s original method while still retaining statistical stationarity. The Reynolds number was varied in the range of Reθ = 670-5500 with shape factor values and skin friction values within 1% and 2%, respectively, of the DNS values given by Orlu et al. (P. Exp Fluids, vol. 54, 2013). Validations were further performed on the higher order moments of velocity profiles as well as energy spectra. Simulations were also run to investigate the effects of stream-wise extent of very-large-scale motions (VLSMs) on boundary layer dynamics. Lastly, a linear stability analysis was conducted on the new stream-wise periodic system to highlight how this new framework can develop insight into boundary layer transition and turbulence. |
Monday, November 19, 2018 9:44AM - 9:57AM |
F21.00009: On the detection of internal interfacial layers in turbulent flows Fan Duosi, Jinglei Xu, Matthew Yao, Jean-Pierre Hickey A novel approach to identify internal interfacial layers, or IILs, in wall-bounded turbulent flows is proposed. Using a Fuzzy Cluster Method (FCM) on the streamwise velocity component, a unique and unambiguous grouping of the Uniform Momentum Zones is achieved, thus allowing the identification of the IILs. The approach overcomes some of the key limitations of the histogram-based IIL identification methods. The method is insensitive to the streamwise domain length, can be used on inhomogeneous grids, uses all the available flow field data, is trivially extended to three dimensions, and does not need user-defined parameters (e.g. number of bins) other than the number of zones, which can be physically justified. The approach is applied to the turbulent boundary layer (experimental, planar PIV) and channel flow (numerical, DNS). The three-dimensional interface identification reveals a streak-like structural organization. The interface is characterized by a strong concentration of spanwise vorticity and is located at the outer edge of the log-layer. |
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