Bulletin of the American Physical Society
66th Annual Meeting of the APS Division of Fluid Dynamics
Volume 58, Number 18
Sunday–Tuesday, November 24–26, 2013; Pittsburgh, Pennsylvania
Session L21: Turbulence: Simulations V |
Hide Abstracts |
Chair: Qiqi Wang, Massachusetts Institute of Technology Room: 316 |
Monday, November 25, 2013 3:35PM - 3:48PM |
L21.00001: Least Squares Shadowing Sensitivity Analysis of Chaotic and Turbulent Fluid Flows Patrick Blonigan, Qiqi Wang, Steven Gomez Computational methods for sensitivity analysis are invaluable tools for fluid dynamics research and engineering design. These methods are used in many applications, including aerodynamic shape optimization and adaptive grid refinement. However, traditional sensitivity analysis methods break down when applied to long-time averaged quantities in chaotic fluid flow fields, such as those obtained using high-fidelity turbulence simulations. This break down is due to the ``Butterfly Effect''; the high sensitivity of chaotic dynamical systems to the initial condition. A new sensitivity analysis method developed by the authors, Least Squares Shadowing (LSS), can compute useful and accurate gradients for quantities of interest in chaotic and turbulent fluid flows. LSS computes gradients using the ``shadow trajectory,'' a phase space trajectory (or solution) for which perturbations to the flow field do not grow exponentially in time. This talk will outline Least Squares Shadowing and demonstrate it on several chaotic and turbulent fluid flows, including homogeneous isotropic turbulence, Rayleigh-B\'{e}nard convection and turbulent channel flow. [Preview Abstract] |
Monday, November 25, 2013 3:48PM - 4:01PM |
L21.00002: Towards Scalable Parallel-in-Time Turbulent Flow Simulations Qiqi Wang, Steven Gomez, Patrick Blonigan, Alastair Gregory, Elizabeth Qian We present a reformulation of unsteady turbulent flow simulations that exhibits chaotic dynamics. Examples include many DNS and LES. This reformulation uses the concept of least squares shadowing. The flow field is assumed to be ergodic, and only long time averaged statistical quantities are considered as quantities of interest. The initial condition is relaxed in the least squares shadowing formulation, and information is allowed to propagate both forward and backward in time. Simulations of chaotic dynamical systems with this reformulation can be proven to be well-conditioned time domain boundary value problems. We analyze how this reformulation can enable scalable parallel-in-time simulation of turbulent flows. [Preview Abstract] |
Monday, November 25, 2013 4:01PM - 4:14PM |
L21.00003: Classification of dense currents over rough walls Raghib Chowdhury, Kiran Bhaganagar Direct numerical simulations and RANS models have been used as a tool to simulate density currents over rough-walls consisting of cylindrical and sinusoidal roughness geometries with different spacing ($\lambda )$ for given height (k) of roughness elements. Scaling laws of front velocity and locations in terms of the spacing between the roughness elements for sinusoidal shaped and sinusoidal roughness element have been obtained. Flow structures for sinusoidal roughness cases revealed that the wake generated at the valley region for sinusoidal or space between the cylinders plays a role on turbulent mixing which leads to reduction in frontal velocity. An important conclusion of the present study is different scaling exist for the k- type and d-type roughness. [Preview Abstract] |
Monday, November 25, 2013 4:14PM - 4:27PM |
L21.00004: FDF in US3D Collin Otis, Pietro Ferrero, Graham Candler, Peyman Givi The scalar filtered mass density function (SFMDF) methodology is implemented into the computer code US3D. This is an unstructured Eulerian finite volume hydrodynamic solver and has proven very effective for simulation of compressible turbulent flows. The resulting SFMDF-US3D code is employed for large eddy simulation (LES) on unstructured meshes. Simulations are conducted of subsonic and supersonic flows under non-reacting and reacting conditions. The consistency and the accuracy of the simulated results are assessed along with appraisal of the overall performance of the methodology. The SFMDF-US3D is now capable of simulating high speed flows in complex configurations. [Preview Abstract] |
Monday, November 25, 2013 4:27PM - 4:40PM |
L21.00005: Three-dimensional simulation of slip-streaming in vehicle aerodynamics Saurav Mitra Simulation of slip-streaming in vehicle aerodynamics is computationally challenging. To resolve turbulent wakes, and estimate drag between two co-linear vehicles with less number of computational cells requires advanced techniques. In this study, the variation of drag reduction and increase arising due to different inter-spacing between two Ahmed vehicles bodies (canonical vehicle geometry with $30^{\circ}$ slant back angle) are presented. The computational fluid dynamics solver CONVERGE was used, for its automatic mesh refinement (AMR) capabilities. AMR is based on the second derivative of shear and normal components of velocity gradients and was used to resolve the flow around geometric features such as the frontal area, the slant back, etc. Steady-state density-based solver is used where each cell has its own pseudo time-step based on the local numerical stability criterion. The RNG k-$\varepsilon $ turbulence model was used to model turbulence. The non-dimensional inter-spacing based on vehicle length, was varied from 0.1 to 2.0. The largest grid size used here was 0.04 m and the smallest was 0.005 m to resolve the turbulent wake which is characterized by a strong vortex system, longitudinal counter-rotating vortices arising from the slant back. [Preview Abstract] |
Monday, November 25, 2013 4:40PM - 4:53PM |
L21.00006: Turbulent transport at rough surfaces Srikanth Toppaladoddi, John Wettlaufer, Sauro Succi We use the Lattice Boltzmann Method to study the effects of rough walls on transport properties at large Reynolds numbers in two dimensions. The roughness elements used have both uniform and non-uniform distributions and we compare our approach with previous studies that have investigated the effects of rough walls on flows in micro channels. The non-uniform roughness distributions have the same spectral properties as that of the underside of Arctic sea ice. [Preview Abstract] |
Monday, November 25, 2013 4:53PM - 5:06PM |
L21.00007: Stochastic field modeling of cavitating flows in OpenFOAM Michael Ranft, Andreas G. Class In [1] analysis is presented for a fluidic diode with low/high pressure drop in forward/reverse flow direction. Accurate description of cavitation is needed due to the dominant effect of vapor bubbles on sound speed. The stochastic field method developed in [1] represents the statistics of growing cavitation bubbles by a set of stochastic fields of vapor fraction which evolve according to the Rayleigh-Plesset equation and local instantaneous LES flow conditions. Cavitation may originate from nucleation sites in the core of turbulent vortices. In this work a RANS model is used instead of LES. Local turbulent pressure fluctuations are recovered based on kinetic energy $k$ of turbulence and its Dissipation $\varepsilon$. In the Rayleigh-Plesset equation these fluctuations are represented by a Wiener process which is superimposed on the mean pressure. Usually a set of stochastic fields is introduced for each stochastic variable. Here two independent Wiener processes, both acting on the vapor-fraction stochastic fields, drive the evolution of vapor bubble growth, so that a single set of stochastic fields can be maintained. The proposed methodology is implemented in OpenFOAM and applied to verification cases including the fluidic diode.\\[4pt] [1] Phys. Fluids 25, 073302 (2013). [Preview Abstract] |
Monday, November 25, 2013 5:06PM - 5:19PM |
L21.00008: Simulating 3D turbulence with Smoothed Particle Hydrodynamics Xiangyu Hu, Stefan Adami, Nikolaus Adams In 2002 Monaghan showed a Lagrangian averaged SPH (Smoothed Particle Hydrodynamics) turbulence model and simulated two-dimensional turbulence. Although achieving good results, this method was shown to be computationally very inefficient (Monaghan, 2002). In this work we present results of 3D turbulence simulated with our newly developed weakly compressible SPH method with modified transport-velocity formulation (Adami, et al., 2013). This fundamental modification was first proposed by Monaghan (Monaghan, 1989). Different from XSPH, we solve a modified momentum equation including a constant background pressure field that regularizes particle motion ``physically'' while strongly reducing artificial numerical dissipation. Numerical results show that the dissipation rate of the 3D Taylor-Green vortex agrees well with DNS results and compared to the standard Smagorinsky model the accuracy is improved (as shown in Figure 1.). To the best knowledge of the authors, this is the first time that a weakly-compress SPH method achieves better results on turbulent flow than the standard grid-based model. [Preview Abstract] |
Monday, November 25, 2013 5:19PM - 5:32PM |
L21.00009: Entropic Lattice Boltzmann Methods for Fluid Mechanics Shyam Chikatamarla, Fabian Boesch, David Sichau, Ilya Karlin With its roots in statistical mechanics and kinetic theory, the lattice Boltzmann method (LBM) is a paradigm-changing innovation, offering for the first time an intrinsically parallel CFD algorithm. Over the past two decades, LBM has achieved numerous results in the field of CFD and is now in a position to challenge state-of-the art CFD techniques. Our major restyling of LBM resulted in an unconditionally stable entropic LBM which restored Second Law (Boltzmann H theorem) in the LBM kinetics and thus enabled affordable direct simulations of fluid turbulence [2]. We review here recent advances in ELBM as a practical, modeling-free tool for simulation of turbulent flows in complex geometries. We shall present recent simulations including turbulent channel flow, flow past a circular cylinder, knotted vortex tubes, and flow past a surface mounted cube. ELBM listed all admissible lattices supporting a discrete entropy function and has classified them in hierarchically increasing order of accuracy[3]. Applications of these higher-order lattices to simulations of turbulence and thermal flows shall also be presented.\\[4pt] [1] Chikatamarla et al, J. Fluid. Mech, 656 (2010); Physica. A, 392 (2013)\\[0pt] [2] Chikatamarla and Karlin, Phys. Rev. Lett., 010201 (2006); Phys. Rev. Lett, 19060 [Preview Abstract] |
Monday, November 25, 2013 5:32PM - 5:45PM |
L21.00010: Efficient error estimation criteria to capture vortical structures in octree meshes Cansu Ozhan, Daniel Fuster, Patrick Da Costa This paper aims at finding optimal adaptive mesh refinement strategies to capture vortical structures. Due to their efficiency, we focus on a-posteriori mesh refinement methods. In particular, we derive a Hessian error estimator for the h-refinement scheme and a residual-based error estimator for finite volume methods and octree grids. The methods are validated for a classical test for the solution of the advection-diffusion-reaction equation and tested against three different test cases where vortical structures are present. In particular we test the temporal evolution of the Lamb-Oseen vortex, the linear growth-rate of small perturbations in a shear viscous layer and the energy evolution in the isotropic turbulence case. The performance of the proposed estimators and the choice of the optimal quantity of interest is discussed for different test cases. [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