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 L28: Particle-laden Flows: Clustering and Droplet Growth |
Hide Abstracts |
Chair: Perrine Pepiot, Cornell University Room: F149 |
Monday, November 21, 2016 4:30PM - 4:43PM |
L28.00001: Numerical investigation of the role of clustering during oxygen-carrier regeneration in Chemical Looping Combustion Himanshu Goyal, Perrine Pepiot In the air-reactor of a dual-bed Chemical Looping Combustion (CLC) system, the spent oxygen-carrier, in the form of metal or reduced metal oxide, is oxidized with air, typically in a high velocity riser reactor. Such a configuration provides challenging modeling issues, as the granular flow is characterized by a highly fluctuating solid volume fraction due to the formation of dense clusters. This may strongly affect the solid residence time in the air-reactor, and therefore, the extent of the oxygen-carrier regeneration and ultimately, the overall reactivity of the carrier in the fuel reactor. Here, we investigate how clustering impacts gas-solid chemical reactions in the reactor using a detailed Lagrange-Euler computational framework. The simulations account for both mass and heat transfer between the gas phase and the metal oxide particles, and the evolution of oxygen content of the metal oxide particles, or equivalently, their degree of oxidation. Two particle models of different complexity are considered. Results are analyzed to quantify the relative importance on the regeneration process of the reactor hydrodynamics. [Preview Abstract] |
Monday, November 21, 2016 4:43PM - 4:56PM |
L28.00002: Transport and Morphology of Nanoparticle Aggregates in Packed Beds: Findings from LBM simulations Ngoc Pham, Dimitrios Papavassiliou Aggregation of colloidal particles in porous media has attracted attention because of possible pore clogging and sedimentation, which reduces the particle breakthrough. In addition, in some systems, further attachment of colloidal particles on deposited aggregates is expected. In this study, the morphology of nanoparticle aggregates, propagating in beds packed with spheres under different electrolyte conditions, is numerically investigated. In our simulation, the nanoparticles are advanced by balancing forces such as drag, random force, buoyancy, gravitational force, electrostatic repulsion, and van der Waals attractive force. When the van der Waals forces take over, the aggregates are formed. The packed beds are made of spheres, either ideally packed or randomly packed in simulation boxes. Sequentially, the flow field of water inside the packed beds is generated, using the lattice Boltzmann method (LBM). In conjunction with that, a Lagrangian framework [1, 2] is applied to record the trajectories of the free nanoparticles and the aggregated nanoparticles. Within the scope of this study, we draw attention to the change of the morphology of the aggregates, reflected by their fractal dimension, under various electrolyte and packing conditions. \textbf{REFERENCES} 1. R. S. Voronov, S. VanGordon, V. I. Sikavitsas, D. V. Papavassiliou, Int. J. Num. Meth. Fluids, 67, 501-517, 2011 2. N.H. Pham, D.P. Swatske, J.H. Harwell, B-J Shiau, D.V. Papavassiliou, Int. J. Heat {\&} Mass Transf., 72, 319-328, 2014. [Preview Abstract] |
Monday, November 21, 2016 4:56PM - 5:09PM |
L28.00003: Columnar structure formation of a dilute suspension of settling spherical particles in a quiescent fluid Sander G. Huisman, Thomas Barois, Mickael Bourgoin, Agathe Chouippe, Todor Doychev, Peter Huck, Carla Bello Morales, Markus Uhlmann, Romain Volk The settling of heavy spherical particles in a column of quiescent fluid is investigated. The performed experiments cover a range of Galileo numbers ($110 \leq \text{Ga} \leq 310$) for a fixed density ratio of $\Gamma = \rho_p/\rho_f = 2.5$. In this regime the wake undergoes several transitions for increasing $\text{Ga}$ resulting in particle motions that are successively: vertical, oblique, oblique oscillating, and finally chaotic. In this work volume fractions up to $\Phi_V = \mathcal{O}\left$ ($10^{-3}\right$) are investigated. Multi-camera recordings of settling particles are recorded and tracked over time in 3 dimensions. A variety of analysis including Vorono\"i analysis and pair angle analysis are performed and show a strong clustering behavior along with an enhancement settling velocity. The experimental findings are compared to simulations. [Preview Abstract] |
Monday, November 21, 2016 5:09PM - 5:22PM |
L28.00004: Effects of clustering on heat transfer in particle-laden turbulence Hadi Pouransari, Ali Mani Particle-laden flows are ubiquitous in variety of natural and industrial phenomena. Rain droplets in clouds, protoplanetary disks, and combustion chambers are examples in which particles are interacting with a background turbulence. It is well known that interaction of particles and turbulent flow results in preferential concentration. The extent of preferential concentration depends on ratio of particle relaxation time and turbulent eddies time scale.this work, we consider particle-laden turbulent flows, in which particles are heated. This is the case for example in the particle-based solar receivers where particles absorb external radiation and heat the background gas. We use three-dimensional variable density direct numerical simulations for the turbulent flow and Lagrangian point-particle tracking to study the implication of particle clustering in particle-to-gas heat transfer. We investigate variety of non-dimensional numbers including particle Stokes number, Reynolds number, and mass loading ratio. Using our statistical analyses we introduce a model to correct the particle-to-gas heat transfer to account for particle clustering. This can be employed in Reynolds average Navier Stokes (RANS) computations. [Preview Abstract] |
Monday, November 21, 2016 5:22PM - 5:35PM |
L28.00005: Coherent clusters of inertial particles in homogeneous turbulence Lucia Baker, Ari Frankel, Ali Mani, Filippo Coletti Clustering of heavy particles in turbulent flows manifests itself in a broad spectrum of physical phenomena, including sediment transport, cloud formation, and spray combustion. However, a clear topological definition of particle cluster has been lacking, limiting our ability to describe their features and dynamics. Here we introduce a definition of coherent cluster based on self-similarity, and apply it to the distribution of heavy particles in direct numerical simulations of homogeneous isotropic turbulence. We consider a range of particle Stokes numbers, with and without the effect of gravity. Clusters show self-similarity at length scales larger than twice the Kolmogorov length, with a specific fractal dimension. In the absence of gravity, clusters demonstrate a tendency to sample regions of the flow where strain is dominant over vorticity, and to align themselves with the local vorticity vector; when gravity is present, the clusters tend to align themselves with gravity, and their fall speed is different from the average settling velocity. This approach yields observations which are consistent with findings obtained from previous studies while opening new avenues for analysis of the topology and evolution of particle clusters in a wealth of applications. [Preview Abstract] |
Monday, November 21, 2016 5:35PM - 5:48PM |
L28.00006: Particle clustering within a two-phase turbulent pipe jet Timothy Lau, Graham Nathan A comprehensive study of the influence of Stokes number on the instantaneous distributions of particles within a well-characterised, two-phase, turbulent pipe jet in a weak co-flow was performed. The experiments utilised particles with a narrow size distribution, resulting in a truly mono-disperse particle-laden jet. The jet Reynolds number, based on the pipe diameter, was in the range $10000\leq Re_{D} \leq40000$, while the exit Stokes number was in the range $0.3\leq Sk_{D} \leq22.4$. The particle mass loading was fixed at $\phi=0.4$, resulting in a flow that was in the two-way coupling regime. Instantaneous particle distributions within a two-dimensional sheet was measured using planar nephelometry while particle clusters were identified and subsequently characterised using an in-house developed technique. The results show that particle clustering is significantly influenced by the exit Stokes number. Particle clustering was found to be significant for $0.3 \leq Sk_{D} \leq 5.6$, with the degree of clustering increasing as $Sk_{D}$ is decreased. The clusters, which typically appeared as filament-like structures with high aspect ratio oriented at oblique angles to the flow, were measured right from the exit plane, suggesting that they were generated inside the pipe. [Preview Abstract] |
Monday, November 21, 2016 5:48PM - 6:01PM |
L28.00007: Global stability of inertial particle trajectories in a two dimensional flow Senbagaraman Sudarsanam, Phanindra Tallapragada The trajectories of inertial particles moving even in a two dimensional fluid flow exhibit complex dynamics, in particular preferential clustering in some sub domains of the fluid. This preferential clustering is influenced by the vorticity field. Based on the Maxey-Riley equation, several Eulerian criteria have been proposed in the past that classify the fluid region into local stable and unstable regions which roughly act as attracting and repelling regions for inertial particles. We demonstrate through examples that the locally unstable regions of the fluid domain can nevertheless act as global attractors. This global stability of unstable regions can partly explain the experimental evidence that particle clustering in fluids is more robust than usually predicted. The example relies on fluid flow generated by point vortices. Such vortex fields are often encountered in several microfluidic flows where the manipulation of the motion of inertial particles has several important applications. [Preview Abstract] |
Monday, November 21, 2016 6:01PM - 6:14PM |
L28.00008: Measurements of the cross-sectional distributions of spherical particles suspended in rectangular channel flows Takahiro Imanishi, Takuya Yabu, Hiroshi Yamashita, Tomoaki Itano, Masako Sugihara-Seki We investigated the inertial migration of neutrally buoyant spherical particles using millimeter-sized rectangular channels of various aspect ratios (\textit{AR} $=$ 1 -- 4.2), in the range of Reynolds numbers (\textit{Re}) from 100 to 2000. The Reynolds number was defined as \textit{UH}/$\nu $, where $U$ is the maximum flow velocity, $H$ is the length of the shorter face of the channel cross-section, and $\nu $ is the kinematic viscosity. Dilute suspensions of polystyrene particles of diameter $d =$ 300 - 650 $\mu $m were used. For the size ratio $d/H \quad =$ 0.1 -- 0.25, the observation of particle positions at downstream cross-sections revealed that the particles were aligned in a straight or curved line nearly parallel to the longer face of the channel cross-section and their probability density function showed a sharp peak at a certain distance from the channel centerline. These focusing positions of particles were found to depend on \textit{Re}, $d/H$ and \textit{AR}. They approached the channel centerline with increasing \textit{Re}. As \textit{AR} increased for constant \textit{Re} and constant $d/H$, focusing positions moved closer to the channel centerline, and reached asymptotic positions for \textit{AR }\textgreater 2. [Preview Abstract] |
Monday, November 21, 2016 6:14PM - 6:27PM |
L28.00009: Correlation of Cloud Droplet Growth with the Scalar Fluctuations in a Turbulent Moist Convection Kamal Kant Chandrakar, Will Cantrell, Kelken Chang, David Ciochetto, Dennis Niedermeier, Mikhail Ovchinnikov, Raymond Shaw, Fan Yang Cloud droplet growth in a turbulent environment is studied by creating turbulent moist Rayleigh-B{\'e}nard convection in the Michigan Tech Pi Chamber. Cloud formation is achieved by injecting aerosols into the water-supersaturated environment created by the isobaric mixing of saturated air at different temperatures. A range of steady-state cloud droplet number concentration is achieved by supplying aerosols at different rates. As steady-state droplet number concentration is decreased the mean droplet size increases as expected, but also the width of the size distribution increases. This increase in the width is associated with larger supersaturation fluctuations due to the slow droplet microphysical response (sink of the water vapor) compared to the fast turbulent mixing (source of the water vapor). The observed standard deviation of the squared droplet radius is a linear function of the combined time scale of the system $\tau_s^{-1} = \tau_c^{-1} + \tau_t^{-1}$; here, $\tau_c$ is the phase relaxation time and $\tau_t$ is the turbulence correlation time. A stochastic differential equation approach for supersaturation also predicts the same linear response. This finding has significance for cloud-radiation budgets and precipitation formation. [Preview Abstract] |
Monday, November 21, 2016 6:27PM - 6:40PM |
L28.00010: Continuous evolution of cloud droplet spectrum in cumulus cloud Toshiyuki Gotoh, Izumi Saito, Takeshi Watanabe We have developed a new method that can seamlessly simulate the continuous growth of cloud droplets to rain drops from the first principle. A cubic box ascending with a mean updraft was introduced and the updraft velocity was self-consistently determined in such a way that the mean turbulent velocity within the box vanished. All the degrees of freedom were numerically integrated by using the Lagrangian dynamics for the droplets and the Eulerian direct numerical simulation for the turbulence. The key processes included were turbulent transport, condensation/evaporation, Reynolds number dependent drag, collision-coalescence, and entrainment. We have examined the evolution of the droplet spectrum over 400 s for a few of the initial droplet spectra: (1) single peak, (2) double peaks, (3) observed distribution, each of which had the same initial mean radius 10$\mu$m and the same mean droplet density $n_p=125$ cm$^{-3}$. The turbulence was in steady state at $R_\lambda=86$ and $\epsilon=33$ cm$^2$s$^{-3}$. It is found that the mass spectrum peak moves slowly toward the larger radius in the early stage and then quickly evolves to have the second peak through the autoconversion to the accretion state. Effects of the condensation and coalescence would also be reported. [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