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 D28: Particle-laden Flows: Simulations |
Hide Abstracts |
Chair: Olivier Desjardins, Cornell University Room: F149 |
Sunday, November 20, 2016 2:57PM - 3:10PM |
D28.00001: A numerical study of bidisperse particles in cluster-induced turbulence Ravi Patel, Bo Kong, Jesse Capecelatro, Rodney Fox, Olivier Desjardins Particle-laden turbulent flow is an important feature of many diverse environmental and industrial systems. To elucidate the mechanics of these types of flows, we study cluster-induced turbulence (CIT), wherein momentum coupling between a carrier fluid and setting particles leads to turbulent-like fluctuations in various quantities of interest. In this work, simulations of CIT with bidisperse particles are presented. The flow of kinetic energy is tracked from its generation due to drag until its dissipation due to fluid viscosity and particle collisions. As suggested by Fox (2014), the particle kinetic energy is separated into a correlated turbulent kinetic energy and an uncorrelated granular energy. An overall energy balance is computed for various exchange terms to determine their relative importance and to understand the underlying physical mechanisms in bidisperse CIT. Additionally, volume fraction and velocity statistics for both particle types and the fluid are presented. From these results, the consequences on closures for Reynolds-averaged stress models of particle-laden flows are discussed. [Preview Abstract] |
Sunday, November 20, 2016 3:10PM - 3:23PM |
D28.00002: Dynamic subgrid-scale modeling for LES of particle-laden turbulent flows Maxime Bassenne, George Ilhwan Park, Javier Urzay, Parviz Moin A new dynamic model is proposed for large-eddy simulations of small inertial particles in turbulent flows. The model is simple, involves no significant computational overhead, and is flexible enough to be deployed in any type of flow solvers and grids, including unstructured setups. The approach does not require any tunable parameters and is based on the use of elliptic differential filters. Particle laden isotropic turbulence and turbulent channel flow are considered. Improved agreement with direct numerical simulation results are observed in the dispersed-phase statistics. The comparisons include analyses of particle acceleration, local carrier-phase velocity, turbophoresis, and preferential-concentration metrics. [Preview Abstract] |
Sunday, November 20, 2016 3:23PM - 3:36PM |
D28.00003: Eulerian-Lagrangian study of particle resuspension by a periodically-forced impinging jet Wen Wu, Giovanni Soligo, Cristian Marchioli, Alfredo Soldati, Ugo Piomelli In this work, we investigate the mechanisms that govern particle resuspension in an impinging flow over surfaces covered with mobile sediments. An Eulerian-Lagrangian approach based on large-eddy simulation of turbulence, and one-way coupling Lagrangian-tracking of particles, is used to model a vertical impinging jet, to which a sequence of periodically-forced azimuthal vortices is superposed. We show how the dynamics of sediments is governed by their interaction with the turbulent structures (including the large-scale vortices) and the separated flow. After initial lift-up from the impingement surface, particles are accumulated in regions where near-wall vortices roll around the impinging azimuthal vortex, forming rib-like structures that either propel particles away from the azimuthal vortex or entrap them in the shear layer between the azimuthal and secondary vortices. We demonstrate that these trapped particles are more likely to reach the outer flow region and generate a persistent cloud of airborne particles. [Preview Abstract] |
Sunday, November 20, 2016 3:36PM - 3:49PM |
D28.00004: A Comparative Study of Euler-Euler and Euler-Lagrange Mesoscale Simulations of Moderately Dense Cluster-induced Gas-Particle Turbulence Bo Kong, Ravi Patel, Jesse Capecelatro, Olivier Desjardins, Rodney Fox Recently Euler-Lagrange (EL) approaches have gained considerable popularity, but the computational cost of resolved EL simulations is often prohibitively high. Therefore, Euler-Euler (EE) approaches, such as kinetic-theory-based two-fluid models (TFM), remain the major workhorse in this area. However, the hydrodynamic assumption in TFM has been proven invalid by many experiments and simulations, especially for dilute particles. Previously, the EE Anisotropic Gaussian approach (EE-AG) has been shown to produce good agreement for key statistic results with EL simulations when particles are dilute. In this work, a novel EE-AG/TFM hybrid solution algorithm, based on different contributions to the particle-phase spatial fluxes, is proposed and implemented in an open-source CFD package. Fully resolved mesoscale simulations of cluster-induced turbulence with moderately dense particles are performed. The detailed comparisons with EL simulations demonstrate that this new EE method can accurately capture the dynamics of the gas-solid flows and produce results comparable to the EL simulations. [Preview Abstract] |
Sunday, November 20, 2016 3:49PM - 4:02PM |
D28.00005: Direct Numerical Simulation of Poly-dispersed Solid-Fluid Systems Erich Essmann, Pei Shui, Rama Govindarajan, Stephane Popinet, Prashant Valluri The fluid dynamics of poly-dispersed solid -- fluid systems are of great importance, particularly is the behaviour of methane clathrates slurries. In this work, a framework is being developed for the direct numerical simulation of these systems. We have extended the Gerris software package of (Popinet et al, 2003). In our solid solver, Gerris Immersed Solid Solver (GISS), to account for collisions we have implemented a novel contact model (Ness {\&} Sun, 2016) for solid-solid interactions. A composite contact model is being used, in which each solid in the domain is divided into two regions. The outer region uses a Hookean repulsive and a lubrication force model to simulate contact. The inner region uses a constraint based contact model to ensure that the numerical overlap of the solids is not excessive. We have validated our methodology against published experimental data. Particularly, we compared the chaotic motion of an ellipsoidal solid in an ideal fluid (Aref, 1993) to that predicted by GISS and the settling behaviour of two colliding spheres of different densities (Zhao, 2003). The validated extensions will allow us to compare previous results from GISS to regimes in which solid-solid contact is important. [Preview Abstract] |
Sunday, November 20, 2016 4:02PM - 4:15PM |
D28.00006: Large-eddy simulation of charged particle flows to model sandstorms Mustafa Rahman, Wan Cheng, Ravi Samtaney Intense electric fields and lightning have been observed in sandstorms. It is proposed to investigate the physical mechanisms essential for production and sustenance of large-scale electric fields in sandstorms. Our central hypothesis is that the turbulent transport of charged sand particles is a necessary condition to attain sustained large-scale electric fields in sandstorms. Our investigation relies on simulating turbulent two-phase (~air and suspended sand particles) flows in which the flow of air is governed by the filtered Navier-Stokes equations with a subgrid-scale model in a Large-Eddy-Simulation setting, while dust particles are modeled using the Eulerian approach using a version of the Direct Quadrature Method of Moments. For the fluid phase, the LES of incompressible turbulent boundary layer employs stretched spiral vortex subgrid-scale model and a virtual wall model similar to the work of Cheng, Pullin \& Samtaney (J. Fluid Mech. 2015). We will quantify the effects of different sand particle distributions, and turbulent intensities on the root-mean-square of the generated electric fields. [Preview Abstract] |
Sunday, November 20, 2016 4:15PM - 4:28PM |
D28.00007: The preferential erosion and deposition of heavy particles over erodible beds Scott Salesky, Marco Giometto, Michael Lehning, Marc Parlange The erosion, transport, and deposition of heavy particles over erodible beds by turbulent flow is a significant process in the context of sediment transport, aeolian processes, and snow transport in alpine and polar regions. While it is well-known that terrain features can lead to spatially inhomogeneous deposition velocities, a systematic study considering the effects of terrain and particle properties has not been conducted to date using large eddy simulation (LES). Using a recently developed Eulerian finite-volume model for the transport of heavy particles over complex terrain in LES, we perform simulations of the transport, erosion, and deposition of heavy particles over idealized surface topography. A new model for particle ejection in the saltation layer subject to the constraints of energy and momentum conservation is adapted for use in an Eulerian framework. A suite of simulations is conducted in order to explore the governing parameters relevant for erosion and deposition (e.g. Stokes number, Rouse number, Shields number, surface cohesion) and to investigate the influence of the mean flow vs. turbulent fluxes for the observed erosion and deposition patterns. Implications for model development will be highlighted, and numerical considerations will be discussed. [Preview Abstract] |
Sunday, November 20, 2016 4:28PM - 4:41PM |
D28.00008: High resolution simulations of down-slope turbidity currents into stratified saline ambient Raphael Ouillon, Senthil Radhakrishnan, Eckart Meiburg, Bruce Sutherland In this work we explore the properties of turbidity currents moving down a slope into a stratified saline ambient through highly resolved 3D Navier-Stokes simulations. Turbidity events are difficult to measure and to replicate experimentally for a wide range of parameters, but they play a key role in ocean, lake or river sediment transport. Our objectives are to improve on previous numerical studies, obtain quantitative data in a more controlled environment than current experimental set-ups, and combine results with analytical arguments to build physics-based scaling laws. We validate our results and propose a simple scaling law to predict the velocity of the front down a slope for any stratification. We also compute a time and space dependent entrainment of ambient fluid and highlight its strong variability. We then introduce a predictable scaling law for the intrusion depth that does not depend on an averaged entrainment and uses it as a verification tool instead. Finally, we show that the ratio of Stokes losses in the local flow around individual particles to dissipative losses of the large scale flow determines the ability of the flow to convert potential energy into kinetic energy. For different parameters, either mechanism can dominate the dynamics of the flow. [Preview Abstract] |
Sunday, November 20, 2016 4:41PM - 4:54PM |
D28.00009: Solids mixing in bubbling fluidized beds: CFD-based analysis of Bubble Dynamics and Time Scales Akhilesh Bakshi, Christos Altantzis, Ahmed Ghoniem In bubbling fluidized bed reactors, solids mixing is critical because it directly affects fuel segregation and residence time. However, there continues to be a lack of understanding because (a) most diagnostic techniques are only feasible in lab-scale setups and (b) the dynamics are sensitive to the operating conditions. Thus, quantitative estimates of mixing (e.g., dispersion coefficient, mixing indices) often span orders of magnitude although it is well accepted that the micro-mixing and gross circulation of solid particles is driven by bubble motion. To quantify this dependence, solids mixing is investigated using fine-grid 3D CFD simulations of a large 50 cm diameter fluidized bed. Detailed diagnostics of the computed flow-field data are performed using MS3DATA, a tool that we developed to detect and track bubbles, and the solids motion is correlated with the spatial and size distribution of bubbles. This study will be useful for quantifying mixing at commercial scales. [Preview Abstract] |
Sunday, November 20, 2016 4:54PM - 5:07PM |
D28.00010: Simulation of Collision of Arbitrary Shape Particles with Wall in a Viscous Fluid. Fazlolah Mohaghegh, H. S. Udaykumar Collision of finite size arbitrary shape particles with wall in a viscous flow is modeled using immersed boundary method. A potential function indicating the distance from the interface is introduced for the particles and the wall. The potential can be defined by using either an analytical expression or level set method. The collision starts when the indicator potentials of the particle and wall are overlapping based on a minimum cut off. A simplified mass spring model is used in order to apply the collision forces. Instead of using a dashpot in order to damp the energy, the spring stiffness is adjusted during the bounce. The results for the case of collision of a falling sphere with the bottom wall agrees well with the experiments. Moreover, it is shown that the results are independent from the minimum collision cut off distance value. Finally, when the particle's shape is ellipsoidal, the rotation of the particle after the collision becomes important and noticeable: At low Stokes number values, the particle almost adheres to the wall in one side and rotates until it reaches the minimum gravitational potential. At high Stokes numbers, the particle bounces and loses the energy until it reaches a situation with low Stokes number. [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