Bulletin of the American Physical Society
69th Annual Meeting of the APS Division of Fluid Dynamics
Volume 61, Number 20
Sunday–Tuesday, November 20–22, 2016; Portland, Oregon
Session M34: Turbulence: DNS Methods, Process, and Analysis |
Hide Abstracts |
Chair: Guillaume Blanquart, Caltech Room: Oregon Ballroom 203 |
Tuesday, November 22, 2016 8:00AM - 8:13AM |
M34.00001: Two-step simulation of velocity and passive scalar mixing at high Schmidt number in turbulent jets K. Jeff Rah, Guillaume Blanquart Simulation of passive scalar in the high Schmidt number turbulent mixing process requires higher computational cost than that of velocity fields, because the scalar is associated with smaller length scales than velocity. Thus, full simulation of both velocity and passive scalar with high Sc for a practical configuration is difficult to perform. In this work, a new approach to simulate velocity and passive scalar mixing at high Sc is suggested to reduce the computational cost. First, the velocity fields are resolved by Large Eddy Simulation (LES). Then, by extracting the velocity information from LES, the scalar inside a moving fluid blob is simulated by Direct Numerical Simulation (DNS). This two-step simulation method is applied to a turbulent jet and provides a new way to examine a scalar mixing process in a practical application with smaller computational cost. [Preview Abstract] |
Tuesday, November 22, 2016 8:13AM - 8:26AM |
M34.00002: From laminar to fully-developed turbulence in a 500 radii long Osborne Reynolds pipe flow: A direct numerical simulation study. Ronald Adrian, Xiaohua Wu, Parviz Moinn We report our new direct numerical simulation results of the Osborne Reynolds' pipe transition problem in a 500 radii long configuration. The inlet disturbance is generated through a three degree narrow wedge. The present radial-mode inlet disturbance is in contrast to our earlier simulation design using a wire-ring at the inlet, which is circumferential-mode in nature (Wu et al, PNAS, 1509451112, 2015). The current mesh size is 16384 x 201 x 512, and the simulation Reynolds number is 6500 based on the pipe diameter and bulk velocity. Statistics in the fully-developed turbulent region are in good agreement with those sampled from an auxiliary short turbulent pipe simulation using the streamwise periodic boundary condition. Frequency spectra of the turbulence kinetic energy are computed at six streamwise stations, namely, 30R, 60R, 90R, 350R, 400R and 450R downstream of the inlet. Surprisingly, spectra in the late transitional region (60R and 90R) exhibit stronger high frequency content than those in the fully-developed turbulent region. Contours of a passive scalar indicate the existence of patchy turbulent spot structures, even in the fully-developed turbulent region. [Preview Abstract] |
Tuesday, November 22, 2016 8:26AM - 8:39AM |
M34.00003: Characteristics of the mixing volume model with the interactions among spatially distributed particles for Lagrangian simulations of turbulent mixing Tomoaki Watanabe, Koji Nagata The mixing volume model (MVM), which is a mixing model for molecular diffusion in Lagrangian simulations of turbulent mixing problems, is proposed based on the interactions among spatially distributed particles in a finite volume. The mixing timescale in the MVM is derived by comparison between the model and the subgrid scale scalar variance equation. A-priori test of the MVM is conducted based on the direct numerical simulations of planar jets. The MVM is shown to predict well the mean effects of the molecular diffusion under various conditions. However, a predicted value of the molecular diffusion term is positively correlated to the exact value in the DNS only when the number of the mixing particles is larger than two. Furthermore, the MVM is tested in the hybrid implicit large-eddy-simulation/Lagrangian-particle-simulation (ILES/LPS). The ILES/LPS with the present mixing model predicts well the decay of the scalar variance in planar jets. [Preview Abstract] |
Tuesday, November 22, 2016 8:39AM - 8:52AM |
M34.00004: Lagrangian study of acceleration in a turbulent channel flow Juan Ignacio Polanco, Ivana Vinkovic, Nickolas Stelzenmuller, Nicolas Mordant A Lagrangian characterisation of a wall-bounded turbulent flow is presented. Tracking of fluid tracers is performed in experiments and direct numerical simulations of channel flow at $\mathit{Re}_\tau \approx 1450$. Near the channel walls, the presence of large-scale streamwise vortices is associated with high-magnitude centripetal accelerations of tracer particles. This presents a clear signature on time autocorrelations of particle acceleration. Temporal cross-correlations and joint probability density functions of acceleration are used to describe the dependency between acceleration components. It is found that, near the walls, negative streamwise accelerations are associated with wall-normal accelerations directed towards the channel centre. This is due to the combined influence of viscous effects and wall-confinement. Good quantitative agreement between experiments and simulations is obtained. Preliminary results on relative dispersion of fluid particle pairs in channel flow are presented and compared with theoretical scalings derived for homogeneous isotropic turbulence. Presented Lagrangian properties can serve as basis for stochastic modelling of transport and dispersion of pollutants by atmospheric flows. [Preview Abstract] |
Tuesday, November 22, 2016 8:52AM - 9:05AM |
M34.00005: Direct numerical simulation of turbulent boundary layer with constant thickness Yichen Yao, Chunxiao Xu, Weixi Huang Direct numerical simulation is performed to turbulent boundary layer (TBL) with constant thickness at $Re_{\theta } =1420.$ Periodic boundary condition is applied in the streamwise direction, and a mean body force equivalent to the convection term in the mean momentum equation is imposed in this direction. The body force is calculated using the published TBL data of Schlatter and Orlu (2010) at $Re_{\theta } =1420$. The presently simulated TBL is compared with the conventional TBL and turbulent channel flow at the prescribed Reynolds number. The turbulent statistics agrees well with that of Schlatter and Orlu (2010). The pre-multiplied energy spectra in current simulation also present high similarity with the conventional TBL, while differ obviously with those in turbulent channel. The successful replication of turbulent boundary in the current simulation provides an alternative method for boundary layer simulation with much less computational cost. Meanwhile, in aspect of both turbulent statistics and flow structures, the current results indicate that the differences between turbulent channel and boundary layer flow mainly caused by the discrepancy in driving force distribution rather than the periodic boundary restriction. [Preview Abstract] |
Tuesday, November 22, 2016 9:05AM - 9:18AM |
M34.00006: Non-Boussinesq effects on buoyancy-driven variable-density turbulence Denis Aslangil, Daniel Livescu, Arindam Banerjee Non-Boussinesq effects on turbulent mixing of a heterogeneous mixture of two incompressible, miscible fluids with different densities are investigated in terms of properly normalized L$^{\mathrm{2m}}$-norms of density gradient by means of high-resolution Direct Numerical Simulations. In a triply periodic three-dimensional domain, the mixing occurs in response to stirring induced by buoyancy-generated motions between two fluids which are initially segregated in random patches. During the flow evolution, the density gradient can reach high values even at low Atwood numbers indicating that non-Boussinesq effects play a crucial role within the flow. The results cover a broad range of Reynolds numbers and non-dimensional density ratios (Atwood numbers, A) including small (A$=$0.05), moderate (A$=$0.25 and 0.5), and high (A$=$0.75) values. An asymmetric behavior is detected on the probability density function of the density gradient at high Atwood numbers. The evolution of the density gradient and the hierarchy of its higher order norms are also investigated by decomposing the flow into the different flow regions by using density as a fluid marker. It is found that the density gradient is much larger in regions of light fluid compared to regions occupied by the heavier fluid, indicating a strong mixing asymmetry between light and heavy fluids. This shows that Boussinesq models may not be adequate even at low density ratios; contrary to what statistics based on the entire domain. [Preview Abstract] |
Tuesday, November 22, 2016 9:18AM - 9:31AM |
M34.00007: Turbulent heat-and-mass transfer in channel flow at transcritical temperature conditions Kukjin Kim, Carlo Scalo, Jean-Pierre Hickey Turbulent heat and mass transfer at transcritical thermodynamic conditions is studied in turbulent channel flow using the high-fidelity DNS for solution to the compressible Navier-Stokes equations in the conservative form closed with the Peng-Robinson state equation. To isolate the real fluid effects on turbulent heat transfer, the bulk pressure is maintained at supercritical $p_b = 1.1 p_c$ and the isothermal walls are set to $\Delta T/2$ above and below the local pseudo-boiling temperature $T_{pb}$ of the fluid (R-134a) where $\Delta T$ is 5K, 10K, and 20K. This setup allows the flow to reach a statistically-steady state while capturing the highest thermodynamic gradients, thus allowing a detailed study on thermodynamics of transcritical turbulent heat transfer. All thermodynamic and turbulent scales are fully resolved which is shown through a careful grid convergence analysis. The time-averaged density and compressibility factor are highly dependent on the temperature field and their large near-wall gradient causes thermodynamically-induced peaks in the RMS quantities resulting in strong turbulent mixing. The ejection of heavy pseudo-liquid blobs by near-wall turbulent structures into the channel core leads to a third RMS peak which is not observable in ideal gas simulations. [Preview Abstract] |
Tuesday, November 22, 2016 9:31AM - 9:44AM |
M34.00008: Quantitative evaluation on contributions of laminar, turbulence and secondary flow to momentum and heat transfer in rhombic ducts Naoya Fukushima In the present study, Direct Numerical Simulation of turbulent flow in rhombic ducts have been carried out to investigate effects of the corner angle on the friction and heat transfer. Due to secondary flow of the second kind, the friction and heat transfer are enhanced in the corner, while turbulence enhances momentum and heat transfer near the wall away from the corner. In previous studies, turbulence and secondary flows are supposed to enhance momentum and heat transfer, qualitatively. The quantitative estimation of their contribution has not been clarified yet. Fukagata, Iwamoto and Kasagi (2002) have theoretically driven the FIK-identity to evaluate quantitative contributions of laminar and turbulence to the friction in turbulent channel. In this study, the FIK-identity has been numerically applied to DNS data in the rhombic ducts to evaluate quantitative contributions of laminar, turbulence and secondary flow to the momentum and heat transfer. From the results, it is quantitatively clarified that the contributions of turbulence and secondary flow to heat transfer are larger than that to friction in the rhombic ducts. [Preview Abstract] |
Tuesday, November 22, 2016 9:44AM - 9:57AM |
M34.00009: Direct numerical simulation of chemical fouling in a turbulent channel flow Hanbyeol Kim, Haecheon Choi In heat exchanger industries, the fouling deposition on solid surfaces causes serious problems such as impaired heat transfer and increased pressure drop. Thus, the prediction and mitigation of fouling deposits has been an important issue. In the present study, we conduct direct numerical simulation of a fully developed channel flow. We assume that fluid flows as a slurry flow and the apparent viscosity is calculated using Thomas' equation. We also assume that a second order reaction with single soluble reactant occurs and a highly viscous product is accumulated at high-temperature wall. Two passive scalar equations are solved along with the Navier-Stokes equations to obtain the mass fraction of the reactant and the temperature in the channel. The reactant flows into the channel at the inlet, and the highly viscous product is stuck on the wall and forms a fouling layer. [Preview Abstract] |
Tuesday, November 22, 2016 9:57AM - 10:10AM |
M34.00010: Computing Normal Shock-Isotropic Turbulence Interaction With Tetrahedral Meshes and the Space-Time CESE Method Balaji Shankar Venkatachari, Chau-Lyan Chang The focus of this study is scale-resolving simulations of the canonical normal shock- isotropic turbulence interaction using unstructured tetrahedral meshes and the space-time conservation element solution element (CESE) method. Despite decades of development in unstructured mesh methods and its potential benefits of ease of mesh generation around complex geometries and mesh adaptation, direct numerical or large-eddy simulations of turbulent flows are predominantly carried out using structured hexahedral meshes. This is due to the lack of consistent multi-dimensional numerical formulations in conventional schemes for unstructured meshes that can resolve multiple physical scales and flow discontinuities simultaneously. The CESE method --- due to its Riemann-solver-free shock capturing capabilities, non-dissipative baseline schemes, and flux conservation in time as well as space --- has the potential to accurately simulate turbulent flows using tetrahedral meshes. As part of the study, various regimes of the shock-turbulence interaction (wrinkled and broken shock regimes) will be investigated along with a study on how adaptive refinement of tetrahedral meshes benefits this problem. [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