Bulletin of the American Physical Society
2005 58th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 20–22, 2005; Chicago, IL
Session FP: Turbulence Simulations: General |
Hide Abstracts |
Chair: Siva Thangam, Stevens Institute of Technology Room: Hilton Chicago Stevens 1 |
Monday, November 21, 2005 8:00AM - 8:13AM |
FP.00001: One-dimensional Turbulence Modeling for a Heated Vertical Wall Harmanjeet Shihn, Paul E. DesJardin In this study, near-wall modeling of heat transfer from a vertically isothermal plate using a one-dimensional modeling (ODT) approach of Kerstein is investigated. The advantage of this approach is that near-wall conduction process can be treated without approximation. The effects of multi-dimensional turbulent mixing processes are modeled using a stochastic process description via triplet mapping stirring events. Adapting the ODT model to the problem of an isothermal plate includes modifying the local characteristic eddy time scale to account for the effects of buoyancy induced mixing mechanisms. Both a Lagrangian and Eulerian implementations of the ODT model are presented. Profiles of time-averaged and RMS velocity and temperature are compared to experimental data and existing self-similarity theories for thermal boundary layer along with Nusselt number predictions. Overall, very good agreement is obtained between simulation results and experimental data including the laminar-to-turbulent transition. [Preview Abstract] |
Monday, November 21, 2005 8:13AM - 8:26AM |
FP.00002: Analysis of ``Poor Man's Navier--Stokes'' and Thermal Energy Equations for High-Rayleigh Number Turbulent Convection J.M. McDonough Derivation of the poor man's Navier--Stokes (PMNS) equations (McDonough \& Huang, {\it Int.\ J.\ Numer.\ Meth.\ Fluids} {\bf 44}, 545, 2004), along with that of the corresponding thermal energy equation, is outlined. These comprise a low-dimensional discrete dynamical system (DDS) that is closely related to the symbol of the differential system and have been shown to be able to efficiently produce any possible Navier--Stokes temporal behavior. Specific relations between the bifurcation parameters of this DDS and the physical dimensionless parameters of the governing equations are presented, and a corresponding approximate expression for the Nusselt number is obtained that can be directly evaluated from the time series of the DDS. Heat transfer correlations are then constructed and shown to compare favorably with results from the low-Prandtl number mercury experiments of Cioni {\it et al., J.\ Fluid Mech.\ }{\bf 335}, 111, 1997 conducted at Rayleigh numbers as high as nearly $10^{10}$. Results suggest potential for use of PMNS equations as temporal part of sub-grid scale model for LES. [Preview Abstract] |
Monday, November 21, 2005 8:26AM - 8:39AM |
FP.00003: Predictive Flow-Field Estimation For Atmospheric Dispersion Problems Mokhasi Paritosh, Dietmar Rempfer In order to address the direct and inverse problems of contaminant dispersion, the need for complete three-dimensional flow-fields at previous and future time steps is very important. The method of Proper Orthogonal Decomposition (POD) allows us to decompose a 3D flow field into a set of basis functions and temporal coefficients. Recently it was shown that it is also possible to reconstruct flow-fields based on the POD and velocity sensor information at a few optimized locations in the domain. Through the use of POD and reduced sensor analysis, it is possible to reconstruct the flow-field at the specific instant of time when the velocity measurements were made. However, the method gives no information about the past or the future behavior of the flow. A new method has been developed that enables one to introduce a pseudo time scale into the POD basis function. This is accomplished by grouping the initial ensemble into subsets of snapshots which we call episodes. When conventional POD analysis is applied to the new ensemble, the resulting basis functions contain the temporal information associated with the flow field also, with the coefficients then varying only in episodes. When coupled with reduced sensor analysis, we obtain a method that enables us to compute the episodic coefficients based on velocity sensors. Once the episodic coefficients are computed, then the velocity field at past, present and future times are known instantly. Numerical experiments conducted on various cases show that the method retains the advantage of POD compression, while still producing accurate results. [Preview Abstract] |
Monday, November 21, 2005 8:39AM - 8:52AM |
FP.00004: An Evaluation of the Flow Simulation Methodology Using the NASA ``Hump'' Geometry Daniel Israel, Hermann Fasel The Flow Simulation Methodology (FSM) is an approach for the simulation of turbulent flows in which a state of the art Reynolds averaged Navier-Stokes (RANS) model is rescaled using a contribution function. This allows the FSM to explicitly resolve only those scales of motion which are well represented on the computational grid while modeling the rest. We present simulations of the separated flow over a wall mounted ``hump'' with unsteady forcing, corresponding to case 3 from the NASA Langley Research Center Workshop: ``CFD Validation of Synthetic Jets and Turbulent Separation Control.'' The turbulence upstream of separation is completely modeled, as in a RANS. In the separated region the FSM performs as a large or very-large eddy simulation. As with most hybrid methods, the FSM is slow in generating fine scale turbulence as it transitions from RANS behavior. However the large structures are well represented and the Reynolds stress distributions show good qualitative agreement with the experiments. The wall pressure fluctuations are relatively easy to capture, and are therefore not a good choice for evaluating turbulence models. Instead we examine the Reynolds stress produced by the phase averaged and the incoherent parts of the resolved motions. This suggests that the principle action of the FSM is to partition energy between the resolved incoherent motions and the model terms, with the strength of the large structures remaining relatively unaffected. [Preview Abstract] |
Monday, November 21, 2005 8:52AM - 9:05AM |
FP.00005: Anisotropic Turbulent Flow Simulations using the Isotropic LANS-$\alpha$ Equations Kamran Mohseni Direct numerical simulation of most engineering and geophysical turbulent flows requires intensive computations. Large Eddy Simulations (LES), Reynolds Averaged Navier-Stokes Equations (RANS), and the Lagrangian averaged Navier-Stokes-$\alpha$ (LANS-$\alpha$) equations are among the numerical techniques to reduce the computational intensity of turbulent flow calculations. In this talk a {\it dynamic} procedure for the Lagrangian Averaged Navier-Stokes-$\alpha$ (LANS-$\alpha$) equations is developed where the variation in the parameter $\alpha$ in the direction of anisotropy is determined in a self-consistent way from data contained in the simulation itself. In order to evaluate the applicability of the dynamic LANS-$\alpha$ model in anisotropic turbulence, {\it a priori} test of the dynamic LANS-$\alpha$ in channel flows is performed at various Taylor Reynolds numbers between 180 and 550 based on the wall friction velocity to find the variation of $\alpha$ in the wall-normal direction. It is found that in the wall region the parameter $\alpha$ rapidly increases away from the wall and saturates to an almost constant value in the outer region. An appropriate scaling for $\alpha$ is also identified. As a result, the isotropic LANS-$\alpha$ equations can now be easily used in anisotropic channel flows with a universally damped $\alpha$. [Preview Abstract] |
Monday, November 21, 2005 9:05AM - 9:18AM |
FP.00006: Modeling vortical flows using linear eddy viscosity closures K. Duraisamy, G. Iaccarino An inherent shortcoming of linear eddy viscosity (LEV) RANS closures is their inability to correctly account for the effects of flow rotation and streamline curvature. Several corrections and modifications have been devised to improve LEV models - these typically involve either a change in the dissipation rate equation or the introduction of coefficients in the turbulence production that depend on the mean velocity gradients. In this work, a modification to LEV models is introduced to improve the predictions of flows dominated by strong vortices. Following the approach of Petterson Reif et al., a correction is devised to mimic the behavior of SMC models in their response to rotation. In particular, a constraint is devised to bound the turbulent production to realizable states in the original SMC closure. Analyses are reported for SMC solutions obtained for homogeneous rotating turbulence and for a free vortex. The correction is used in combination with the $v^2-f$ turbulence model, although, in general, it can be applied to any of the conventional LEV models. An example of its application to a 1 eqn. LEV model is presented. An assessment of the proposed modification on modeling trailing vortex flow-fields is performed. Reference data include DNS of an isolated axisymmetric vortex at a moderate vortex Reynolds number (Re) ($\approx 16500$) and experimental measurements in the wake of a wing at a high Re ($\approx 2 \times 10^6$). The level of agreement achieved with the reference data confirms the viability of the approach. [Preview Abstract] |
Monday, November 21, 2005 9:18AM - 9:31AM |
FP.00007: On the behavior of two-equation turbulence models in the limit of low turbulence levels: A dynamical systems approach B. Anders Pettersson Reif, Chris Rumsey, Thomas Gatski The utilization of the Reynolds-averaged Navier-Stokes (RANS) approach continues to dominate computational predictions of turbulent flows in a wide range of disciplines. Despite the simplistic statistical treatment of the turbulence motion, RANS prevails as the primary computational tool in the vast majority of applications. The present study is motivated by the challenges posed in conjunction with computing transitional flows, in particular by the need to better understand the response of models that routinely are applied in complex turbulent flow computations where regions of very low turbulence levels exist. A dynamical systems analysis is employed in this study to shed light on the dynamical behavior of the most commonly used two-equation models, and to aid the identification of inherent limitations that may have practical consequences. It is demonstrated that some common forms of the K-e model can yield arbitrary steady state solutions in transitional flows that depend on numerical solution parameters such as initial conditions and solution methodologies. In particular models that utilizes a wall damping coefficient in the destruction term of the dissipation rate transport equation. A so-called null-cline analysis will be introduced as a useful tool to analyze the solution of models equations near critical points. [Preview Abstract] |
Monday, November 21, 2005 9:31AM - 9:44AM |
FP.00008: Experimental and Computational Studies of Wake flows Siva Thangam, Igbal Mehmedagic, Donald Carluccci Experiments are performed in low-speed wind tunnels to analyze flow past sting-mounted cylinders. Rear-mounted and fore-mounted stings are utilized to perform experiments. Computations are performed using a two-equation turbulence model that is capable of capturing the effects of swirl and curvature. The model performance was validated with benchmark experimental flows and implemented for analyzing the flow configuration used in the experimental study. The Reynolds number range of 260000 and rotation numbers of up to 1.2 (based on cylinder diameter) are considered for both stationary cylinders and those with a rotating base. The results are analyzed and the predictive capability of the model for flows with swirl is discussed. [Preview Abstract] |
Monday, November 21, 2005 9:44AM - 9:57AM |
FP.00009: Elliptic instability in the alpha-model family of turbulence models in MHD Bruce Fabijonas, Darryl Holm We examine elliptic instability in the MHD analogue of the family of NS turbulence models known as alpha-models. The instability is an exact solution of the classical MHD equations as well as the discussed turbulence models for MHD. The instability arises when a wave whose wave vector has general temporal behavior traveling in three dimensions interacts with a two dimensional swirling flow. Such waves are known as Kelvin waves. We use the instability to study how various turbulence models affect the classical solutions of the MHD equations. [Preview Abstract] |
Monday, November 21, 2005 9:57AM - 10:10AM |
FP.00010: Dynamo in the Taylor-Green vortex: Direct numerical simulations and modeling of MHD flows Annick Pouquet, Pablo Mininni, David Montgomery, Jean-Francois Pinton, Helene Politano, Yannick Ponty Direct numerical simulations (DNS) and Lagragian-averaged model runs (LAMHD) of three-dimensional magnetohydrodynamic turbulence are presented. The model allows for a significant reduction of computer resources at given Reynolds numbers. It correctly reproduces the growth rate of magnetic energy and captures the nonlinear saturation level; intermittency is recovered as well. Low magnetic Prandlt number dynamos are then explored combining DNS, LAMHD and Large-Eddy Simulations. The flow is forced with a Taylor-Green non-helical vortex with a well-defined structure at large scales and strong turbulent fluctuations. Dynamos are observed down to the lowest PM=0.01 that can be modeled accurately for this flow; the critical magnetic Reynolds number increases sharply with PM as turbulence sets in and then saturates; in the linear phase, the most unstable magnetic modes move to small scales as PM is decreased; a Kazantsev 3/2 spectrum develops with strong non-local nonlinear transfer; then the dynamo grows at large scales and modifies the turbulent velocity fluctuations. Other forcing including Beltrami flows are found to behave in a similar fashion. [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