Bulletin of the American Physical Society
68th Annual Meeting of the APS Division of Fluid Dynamics
Volume 60, Number 21
Sunday–Tuesday, November 22–24, 2015; Boston, Massachusetts
Session L9: CFD: Large Eddy Simulation I |
Hide Abstracts |
Chair: Matthias Ihme, Stanford University Room: 109 |
Monday, November 23, 2015 4:05PM - 4:18PM |
L9.00001: LES-Modeling of a Partially Premixed Flame using a Deconvolution Turbulence Closure Qing Wang, Hao Wu, Matthias Ihme The modeling of the turbulence/chemistry interaction in partially premixed and multi-stream combustion remains an outstanding issue. By extending a recently developed constrained minimum mean-square error deconvolution (CMMSED) method, to objective of this work is to develop a source-term closure for turbulent multi-stream combustion. In this method, the chemical source term is obtained from a three-stream flamelet model, and CMMSED is used as closure model, thereby eliminating the need for presumed PDF-modeling. The model is applied to LES of a piloted turbulent jet flame with inhomogeneous inlets, and simulation results are compared with experiments. Comparisons with presumed PDF-methods are performed, and issues regarding resolution and conservation of the CMMSED method are examined. [Preview Abstract] |
Monday, November 23, 2015 4:18PM - 4:31PM |
L9.00002: Scale-truncating relaxation models for large eddy simulations Roel Verstappen, Maurits Silvis This paper discusses novel relaxation models for large eddy simulation (LES) of turbulent flows. To verify that the scales of motion are truncated properly by the LES-model an explicit box filter is introduced. The relaxation parameter is then determined such that the production of all box-fitting scales is counterbalanced by the dissipation associated with the relaxation model. This balance is imposed at the discrete level; here using a second-order finite-volume discretization. Notice that the approach can in principle be applied to any discretization method. The resulting relaxation parameter depends on the invariants of the discrete velocity gradient. The model is successfully tested for canonical turbulent flows (isotropic turbulence, turbulent channel flow, mixing layer). [Preview Abstract] |
Monday, November 23, 2015 4:31PM - 4:44PM |
L9.00003: ABSTRACT WITHDRAWN |
Monday, November 23, 2015 4:44PM - 4:57PM |
L9.00004: LES of propelled bodies in crashback Praveen Kumar, Krishnan Mahesh Crashback is an off-design operating condition to quickly stop a propelled vehicle by rotating the propeller in reverse direction, thus yielding a negative thrust. The interaction of the freestream with the strong reverse flow from the propeller creates massive unsteadiness and flow separation. This talk will discuss our work towards simulation of crashback flow over an entire hull using Large-Eddy Simulation (LES).The results will be compared to the available experimental data and the flow physics will be discussed. The flowfield of the hull-attached propeller in crashback will be analyzed using dynamic mode decomposition to understand the mechanism of the unsteady loads. [Preview Abstract] |
Monday, November 23, 2015 4:57PM - 5:10PM |
L9.00005: Large eddy simulation on unstructured meshes using Lagrangian subgrid-scale model for complex turbulent flows Steven Tran, Onkar Sahni Large eddy simulations (LES) provide high fidelity in which the large-scale turbulent structures are resolved while their interactions with the subgrid scales are modeled. In a Smagorinsky-based LES approach, the unresolved stresses are modeled using an eddy viscosity which in-turn involves a model parameter that is unknown \textit{a priori} and varies in space and time for complex problems. Therefore, dynamic procedures are employed to determine this parameter where averaging is applied to make the procedure robust. When applicable, spatial averaging is applied across homogeneous directions. However, for complex flows the Lagrangian subgrid-scale model employing averaging over pathlines becomes attractive. In contrast to the dynamic Smagorinsky model, variational multiscale (VMS) models have also been developed for LES. \ In this study, we investigate dynamic mixed models for LES based on the combinations of the Lagrangian subgrid-scale model and the residual-based VMS (RBVMS) approach to study complex, inhomogeneous turbulent flows on unstructured meshes. Applications range from flow through a channel to flow over an airfoil at a moderate angle of attack. Experimental and DNS data are used to make comparisons. [Preview Abstract] |
Monday, November 23, 2015 5:10PM - 5:23PM |
L9.00006: Large eddy simulation on deforming unstructured meshes using Lagrangian subgrid-scale model Reed Cummings, Onkar Sahni, Steven Tran Large eddy simulation (LES) provides a high fidelity alternative to direct numerical simulation (DNS) in which computational costs are significantly reduced. LES resolves the large-scale turbulent structures present in the flow while modeling the effect of the unresolved or subgrid scales on resolved scales. The Smagorinksy-based LES approach uses an eddy viscosity to model the unresolved stresses. These models rely on a parameter to calculate the eddy viscosity. For complex flows this parameter varies with space and time, thus dynamic procedures are required to determine the parameter. Additionally, many flows problems of interest involve moving geometries or deforming domains for which appropriate dynamic procedure is required. This work employs the Lagrangian subgrid-scale model within an arbitrary Lagrangian Eulerian (ALE) formulation involving deforming unstructured meshes. Two cases will be studied including flow through a channel and flow over an airfoil. [Preview Abstract] |
Monday, November 23, 2015 5:23PM - 5:36PM |
L9.00007: Numerical investigation of the convective heat transfer coefficient with longitudinal pitch variation in a staggered tube bank Ashraf Alfandi, Juhyeon Yoon, Khalifeh Abusaleem, Mohammad Albati, Salih Khafaji In this study, the effect on a shell-side heat transfer coefficient is investigated using the CFD code FLUENT with a variation in longitudinal pitch to diameter ratio, SL, in the range of 1.15 to 2.6 with a fixed transverse pitch to diameter ratio. For the benchmark purposes with the available empirical correlation, typical thermal-hydraulic conditions for the Zukauskas correlation are assumed. Many sensitivity calculations for different mesh sizes and turbulent models are performed to check the accuracy of the numerical solution. A realizable $\kappa $-$\varepsilon $ turbulence model was found to be in good agreement with results of the Zukauskas correlation among the other turbulence models, at least for the staggered tube bank. It was found that the average heat transfer coefficient of a crossflow over a staggered tube bank calculated using FLUENT is in good agreement with the Zukauskas correlation-calculated heat transfer coefficient in the range of 1.15 -- 2.6. For a staggered tube bank, using the Zukauskas correlation seems to be valid down to SL $=$ 1.15. [Preview Abstract] |
Monday, November 23, 2015 5:36PM - 5:49PM |
L9.00008: Implicit Large Eddy Simulation of a wingtip vortex at Re$_{\mathrm{c}}$ $=$1.2x10$^{6}$ Jean-Eloi Lombard, Dave Moxey, Spencer Sherwin We present recent developments in numerical methods for performing a Large Eddy Simulation (LES) of the formation and evolution of a wingtip vortex. The development of these vortices in the near wake, in combination with the large Reynolds numbers present in these cases, make these types of test cases particularly challenging to investigate numerically. To demonstrate the method's viability, we present results from numerical simulations of flow over a NACA 0012 profile wingtip at Re$_{\mathrm{c}} = $ 1.2 x10$^{6}$ and compare them against experimental data, which is to date the highest Reynolds number achieved for a LES that has been correlated with experiments for this test case. Our model correlates favorably with experiment, both for the characteristic jetting in the primary vortex and pressure distribution on the wing surface. The proposed method is of general interest for the modeling of transitioning vortex dominated flows over complex geometries. [Preview Abstract] |
Monday, November 23, 2015 5:49PM - 6:02PM |
L9.00009: Large-eddy simulations of a propelled submarine model Antonio Posa, Elias Balaras The influence of the propeller on the wake as well as the evolution of the turbulent boundary layers over an appended notional submarine geometry (DARPA SUBOFF) is reported. The present approach utilizes a wall-resolved LES, coupled with an immersed boundary formulation, to simulate the flow model scale Reynolds numbers ($Re=1.2e+06$, based on the free-stream velocity and the length of the body). Cylindrical coordinates are adopted, and the computational grid is composed of 3.5 billion nodes. Our approach has been validated on the appended submarine body in towed conditions (without propeller), by comparisons to wind tunnel experiments in the literature. The comparison with the towed configuration shows profound modifications in the boundary layer over the stern surface, due to flow acceleration, with higher values of turbulent kinetic energy in the inner layer and lower values in the outer layer. This behavior was found tied to a different topology of the coherent structures between propelled and towed cases. The wake is also highly affected, and the momentum deficit displays a non-monotonic evolution downstream. An axial peak of turbulent kinetic energy replaces the bimodal distribution of the stresses in the wake, observed in the towed configuration. [Preview Abstract] |
Monday, November 23, 2015 6:02PM - 6:15PM |
L9.00010: Numerical Analysis of the Acoustic Field of Tip-Clearance Flow S.M. Alavi Moghadam Numerical simulations of the acoustic field generated by a shrouded axial fan are studied by a hybrid fluid-dynamics-acoustics method. In a first step, large-eddy simulations are performed to investigate the dynamics of tip clearance flow for various tip gap sizes and to determine the acoustic sources. The simulations are performed for a single blade out of five blades with periodic boundary conditions in the circumferential direction on a multi-block structured mesh with 1.4$\times10^8$ grid points. The turbulent flow is simulated at a Reynolds number of 9.36$\times10^5$ at undisturbed inflow condition and the results are compared with experimental data. The diameter and strength of the tip vortex increase with the tip gap size, while simultaneously the efficiency of the fan decreases. In a second step, the acoustic field on the near field is determined by solving the acoustic perturbation equations (APE) on a mesh for a single blade consisting of approx. 9.8$\times10^8$ grid points. The overall agreement of the pressure spectrum and its directivity with measurements confirm the correct identification of the sound sources and accurate prediction of the acoustic duct propagation. The results show that the longer the tip gap size the higher the broadband noise level. [Preview Abstract] |
Monday, November 23, 2015 6:15PM - 6:28PM |
L9.00011: New approaches to the design optimization of hydrofoils Pooriya Beyhaghi, Gianluca Meneghello, Thomas Bewley Two simulation-based approaches are developed to optimize the design of hydrofoils for foiling catamarans, with the objective of maximizing efficiency (lift/drag). In the first, a simple hydrofoil model based on the vortex-lattice method is coupled with a hybrid global and local optimization algorithm that combines our Delaunay-based optimization algorithm with a Generalized Pattern Search. This optimization procedure is compared with the classical Newton-based optimization method. The accuracy of the vortex-lattice simulation of the optimized design is compared with a more accurate and computationally expensive LES-based simulation.~ In the second approach, the (expensive) LES model of the flow is used directly during the optimization. A modified Delaunay-based optimization algorithm is used to maximize the efficiency of the optimization, which measures a finite-time averaged approximation of the infinite-time averaged value of an ergodic and stationary process. Since the optimization algorithm takes into account the uncertainty of the finite-time averaged approximation of the infinite-time averaged statistic of interest, the total computational time of the optimization algorithm is significantly reduced. Results from the two different approaches are compared. [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