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 Q29: Turbulence: Modeling - GeneralTurbulence
|
Hide Abstracts |
Chair: Zhen-Su She, Peking University Room: 205 |
Tuesday, November 21, 2017 12:50PM - 1:03PM |
Q29.00001: PEVC-FMDF for Large Eddy Simulation of Compressible Turbulent Flows Arash Nouri Gheimassi, Mehdi Nik, Peyman Givi, Daniel Livescu, Stephen Pope The filtered density function (FDF) closure is extended to a ``self-contained'' format to include the subgrid scale (SGS) statistics of all of the hydro-thermo-chemical variables in turbulent flows. These are the thermodynamic pressure, the specific internal energy, the velocity vector, and the composition field. In this format, the model is comprehensive and facilitates large eddy simulation (LES) of flows at both low and high compressibility levels. A transport equation is developed for the joint ``pressure-energy-velocity-composition filtered mass density function (PEVC-FMDF).'' In this equation, the effect of convection appears in closed form. The coupling of the hydrodynamics and thermochemistry is modeled via a set of stochastic differential equation (SDE) for each of the transport variables. This yields a self-contained SGS closure. For demonstration, LES is conducted of a turbulent shear flow with transport of a passive scalar. The consistency of the PEVC-FMDF formulation is established, and its overall predictive capability is appraised via comparison with direct numerical simulation (DNS) data. [Preview Abstract] |
Tuesday, November 21, 2017 1:03PM - 1:16PM |
Q29.00002: Scale-Resolving simulations (SRS): How much resolution do we really need? Filipe M.S. Pereira, Sharath Girimaji Scale-resolving simulations (SRS) are emerging as the computational approach of choice for many engineering flows with coherent structures. The SRS methods seek to resolve only the most important features of the coherent structures and model the remainder of the flow field with canonical closures. With reference to a typical Large-Eddy Simulation (LES), practical SRS methods aim to resolve a considerably narrower range of scales (reduced physical resolution) to achieve an adequate degree of accuracy at reasonable computational effort. While the objective of SRS is well-founded, the criteria for establishing the optimal degree of resolution required to achieve an acceptable level of accuracy are not clear. This study considers the canonical case of the flow around a circular cylinder to address the issue of `optimal’ resolution. Two important criteria are developed. The first condition addresses the issue of adequate resolution of the flow field. The second guideline provides an assessment of whether the modeled field is canonical (stochastic) turbulence amenable to closure-based computations. [Preview Abstract] |
Tuesday, November 21, 2017 1:16PM - 1:29PM |
Q29.00003: An investigation of implicit turbulence modeling for laminar-turbulent transition in natural convection ChungGang Li, Makoto Tsubokura, WeiHsiang Wang The automatic dissipation adjustment (ADA) model [J. Comput. Phys. 345 (2017) 462-474] based on truncated Navier-Stokes equations is utilized to investigate the feasibility of using implicit large eddy simulation (ILES) with ADA model on the transition in natural convection. Due to the high Rayleigh number coming from the larger temperature difference (300K), Roe scheme modified for low Mach numbers coordinating ADA model is used to resolve the complicated flow field. Based on the qualitative agreement of the comparisons with DNS and experimental results and the capability of numerically predicating a -3 decay law for the temporal power spectrum of the temperature fluctuation, this study thus validates the feasibility of ILES with ADA model on turbulent natural convection. With the advantages of ease of implementation because no explicit modeling terms are needed and nearly free of tuning parameters, ADA model offers to become a promising tool for turbulent thermal convection. [Preview Abstract] |
(Author Not Attending)
|
Q29.00004: A predictive universal fractional-order differential model of wall-turbulence Fangying Song, George Karniadakis Fractional calculus has been around for centuries but its use in computational since and engineering has emerged only recently. Here we develop a relatively simple one-dimensional model for fully-developed wall-turbulence that involves a fractional operator with variable fractional order. We use available DNS data bases to ``learn" the function that describes the fractional order, which has a high value at the wall and decays monotonically to an asymptotic value at the centerline. We show that this function is universal upon re-scaling and hence it can be used to predict the mean velocity profile at all Reynolds numbers. We demonstrate the accuracy of our universal fractional model for channel flow at high Reynolds number as well as for pipe flow and we obtain good agreement with the Princeton super-pipe data up to Reynolds numbers 35,000,000. [Preview Abstract] |
Tuesday, November 21, 2017 1:42PM - 1:55PM |
Q29.00005: Turbulence modeling with fractional derivatives: Derivation from first principles and initial results Brenden Epps, Benoit Cushman-Roisin Fluid turbulence is an outstanding unsolved problem in classical physics, despite 120+ years of sustained effort. Given this history, we assert that a new mathematical framework is needed to make a transformative breakthrough. This talk offers one such framework, based upon kinetic theory tied to the statistics of turbulent transport. Starting from the Boltzmann equation and “Lévy α-stable distributions”, we derive a turbulence model that expresses the turbulent stresses in the form of a fractional derivative, where the fractional order is tied to the transport behavior of the flow. Initial results are presented herein, for the cases of Couette-Poiseuille flow and 2D boundary layers. Among other results, our model is able to reproduce the logarithmic Law of the Wall in shear turbulence. [Preview Abstract] |
Tuesday, November 21, 2017 1:55PM - 2:08PM |
Q29.00006: A near-wall model from analysis of inner-outer interactions in filtered wall bounded 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. We show that in this limit the filtered velocities $\bar{u}$ 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 $f$ in the near-wall region. The profile $f$ is evaluated through analysis of DNS data from channel flow at $Re_{\tau}=5200$. We use the resulting profile $f$ to formulate a predictive near-wall model. The model depends only on the filtered velocities at the boundary of the near-wall layer and can supply boundary conditions for a wall-modeled LES. We present preliminary results from a coarse LES using this wall model. [Preview Abstract] |
Tuesday, November 21, 2017 2:08PM - 2:21PM |
Q29.00007: Turbulence closures in variable density jet flow John Charonko, Kathy Prestridge Variable density mixing plays a role in a variety of physical systems across many orders of length and time scales. Frequently, these effects take place in buoyant jets and plumes. While turbulence has been studied in these systems for many years, theory is less developed than for single-fluid jets and when modeling these systems using density-weighted (Favre) Reynolds-averaged equations it is not clear which closures developed for constant-density flows also apply to cases with large density ratios. Here we present an analysis of several turbulence closures compared to the exact forms of the correlations using a previously acquired dataset (Charonko and Prestridge 2017) from an open circuit wind tunnel featuring initial buoyant jet conditions of Re $\approx$ 19,000 and $\rho_{jet}/\rho_\infty=1.2$ and $4.2$, measured with simultaneous PIV and acetone PLIF. This allows computation of the correlated velocity and density terms in the turbulent kinetic energy and Reynolds stress budgets. In particular we will focus on the turbulent transport, including the velocity triple correlation and the pressure-velocity terms, and evaluate several proposed models by comparing the directly-calculated dissipation rate to the residual from the proposed models and the remaining budget terms. [Preview Abstract] |
Tuesday, November 21, 2017 2:21PM - 2:34PM |
Q29.00008: Variable Density Turbulence Spectra using the Sparse-Direct Interaction Perturbation David Petty, Carlos Pantano Equations for the turbulence kinetic energy, velocity-scalar, and scalar power spectra have been derived for homogeneous isotropic variable-density turbulence. A perturbation about the incompressible limit is used to determine the leading order variable-density contribution to the two-point statistical quantities. Closure is achieved using the Sparse Direct-Interaction Perturbation (SDIP) technique. The resulting integro-differential equations are solved using a custom developed numerical method which leverages the Automatic Differentiation by OverLoading in C++ (ADOL-C) library. The numerical solutions indicate only minor changes to the velocity fields due to turbulence-scalar coupling. However, dilatation production is shown to generate turbulence scalar flux divergence even under statistically isotropic conditions. In addition, the scalar power spectrum can undergo anti-correlation under certain conditions. These model spectra are compared to those calculated from Direct Numerical Simulation of the same flow fields. [Preview Abstract] |
Tuesday, November 21, 2017 2:34PM - 2:47PM |
Q29.00009: A Taylor-Microscale Transport Model for RANS Daniel Israel, Abigail Hsu, Joshua Rudolph Since the first development of complete two-equation RANS models, there have been a variety of proposals for the choice of a second scaling quantity. Some of the most popular have been a time-scale $\omega$ (Kolmogorov, 1942; Wilcox, 1998), the dissipation rate $\varepsilon$ (Harlow et al., 1968), the integral length-scale $L$, and the product $kl$ (Mellor et al., 1982). All of these are formally equivalent in the production and dissipation terms, and differ only in which quantity is turbulently diffused. They also all rely on an equilibrium assumption that links the dissipation rate at the small scales to the scale of the large eddies. We propose using the Taylor microscale as the second scale. This has several nice properties, and also exhibits some interesting mathematical differences from conventional models. We show results for some simple shear flows using the new model. [Preview Abstract] |
Tuesday, November 21, 2017 2:47PM - 3:00PM |
Q29.00010: ABSTRACT WITHDRAWN |
Tuesday, November 21, 2017 3:00PM - 3:13PM |
Q29.00011: A new algebraic turbulence model for accurate description of airfoil flows. Meng-Juan Xiao, Zhen-Su She We report a new algebraic turbulence model (SED-SL) based on the SED theory, a symmetry-based approach to quantifying wall turbulence. The model specifies a multi-layer profile of a stress length (SL) function in both the streamwise and wall-normal directions, which thus define the eddy viscosity in the RANS equation (e.g. a zero-equation model). After a successful simulation of flat plate flow (APS meeting, 2016), we report here further applications of the model to the flow around airfoil, with significant improvement of the prediction accuracy of the lift (C$_{\mathrm{L}})$ and drag (C$_{\mathrm{D}})$ coefficients compared to other popular models (e.g. BL, SA, etc.). Two airfoils, namely RAE2822 airfoil and NACA0012 airfoil, are computed for over 50 cases. The results are compared to experimental data from AGARD report, which shows deviations of C$_{\mathrm{L\thinspace }}$bounded within 2{\%}, and C$_{\mathrm{D\thinspace }}$within 2 counts (10$^{\mathrm{-4}})$ for RAE2822 and 6 counts for NACA0012 respectively (under a systematic adjustment of the flow conditions). In all these calculations, only one parameter (proportional to the Karmen constant) shows slight variation with Mach number. The most remarkable outcome is, for the first time, the accurate prediction of the drag coefficient. The other interesting outcome is the physical interpretation of the multi-layer parameters: they specify the corresponding multi-layer structure of turbulent boundary layer; when used together with simulation data, the SED-SL enables one to extract physical information from empirical data, and to understand the variation of the turbulent boundary layer. [Preview Abstract] |
Tuesday, November 21, 2017 3:13PM - 3:26PM |
Q29.00012: SED-based theoretical determination of empirical coefficients in an engineering turbulence model. Fan Tang, Zhen-Su She Based on a recent symmetry-based SED theory of wall turbulence, we derive a series of empirical coefficients used in a popular engineering {\$}k-$\backslash $omega{\$} model for simulating turbulent pipe flow. The SED theory proposes multi-layer expressions for two length (order) functions (e.g. a stress-length and an energy length), which determine the distribution of the Reynolds stress and kinetic energy in the wall-normal direction. Three local solutions, corresponding to viscous sublayer, overlap region and central core layer, respectively, are derived, and empirical model coefficients are determined in terms of universal physical constants of wall turbulence, including the thicknesses of the viscous sublayer, buffer layer and central core layer, as well as a centerline kinetic energy. This derivation rectifies a mistake of the wall scaling in the current {\$}k-$\backslash $omega{\$} model. The most interesting outcome is the prediction of an anomalous scaling in a meso-layer as a part of the overlap region. This derivation yields an explanation why {\$}k-$\backslash $omega{\$} model gives reasonable predictions for industrial flows. [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