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 G3: Particle-Laden Flows: Clustering and Dispersion I |
Hide Abstracts |
Chair: Lance Collins, Cornell University Room: 102 |
Monday, November 23, 2015 8:00AM - 8:13AM |
G3.00001: Inertial Particle Relative Velocity in a High-Reynolds-Number Homogeneous and Isotropic Turbulence Chamber Zhongwang Dou, Zachary Pecenak, Zach Liang, Lujie Cao, Peter Ireland, Lance Collins, Hui Meng Particle-pair radial relative velocity (RV) in turbulence plays a critical role in droplet collision and cloud formation. Both simulations and experiments are performed to better understand RV of inertial particles in homogeneous and isotropic turbulence (HIT). However, past experimental measurement of particle RV statistics exhibited large deviations from DNS results (de Jong et al., 2010). In the current study, we identified intrinsic limitations in our previous study and devised a 4-frame particle tracking velocimetry technique to measure particle RV. In a second-generation, enclosed, fan-driven HIT chamber, both tracer and inertial particles were studied at R\textunderscore $\lambda $ of 366. The experimentally measured RV statistics were compared with DNS with excellent agreement. Additionally, for both kinds of particles, the mean inward RV vs. particle separation distance $r$ also matched very well with DNS, but at near-zero $r$, experimental values were slightly higher. To investigate the cause of this discrepancy, we compared DNS of both mono- and tri-dispersed particles. We found that the tri-dispersed particles exhibited higher mean inward RV at small $r$ than any mono-dispersed particles. This suggests that the increase of mean inward RV in the experiment could be due to the Stokes number (\textit{St}) distribution present in the particles, while DNS employed single \textit{St} values. [Preview Abstract] |
Monday, November 23, 2015 8:13AM - 8:26AM |
G3.00002: Turbulent particle clustering in a fully developed square channel flow. Laura Villafane, Andrew Banko, Chris Elkins, John Eaton Particle-turbulence interactions are investigated in a fully developed turbulent channel air flow to determine the gas phase effect on the particle concentration and velocity fields. The experimental apparatus is a vertical channel with square cross section. The Reynolds number based on channel width is about 10$^{\mathrm{4}}$. The 12 um nominal diameter nickel particles are smaller than the Kolmogorov length scale and the corresponding Stokes number is of the order of one. Low volume and mass loading ratios are considered. Under these conditions preferential concentration is expected to be strong while the effect of particles on the gas flow negligible. The square channel flow contains mean secondary flows not present in high aspect ratio channels studied previously. These will increase transport of particles away from the walls and raise turbulence levels in the central region. Current experiments are focused on the statistics of the particle phase including particle concentration inhomogeneities and particle velocities. The particle concentration field is analyzed from high resolution laser illuminated planar images. Particle velocity distributions are measured by PIV techniques and compared to the carrier-phase mean velocities from aerodynamic pressure measurements in the unladen case. The effect of increasing mass loading ratio on the particle velocities is analyzed. [Preview Abstract] |
Monday, November 23, 2015 8:26AM - 8:39AM |
G3.00003: An asymptotic analysis of particle clustering in turbulent flows Mahdi Esmaily Moghadam, Ali Mani Interaction of dense inertial particles with turbulent flow is analysed. An asymptotic solution is obtained that quantifies particle clustering on a wide range of Stokes numbers and flow conditions. In a simplified form, particle clustering is predicted to be $St/(St^2 + 1)$, in which $St$ is the Stokes number based on the Kolmogorov time scale, hence predicting maximum clustering at $St=1$ and first order decay of clustering as $St \to 0$ and $\infty$. These results are validated against numerical simulation of inertial particles in a homogeneous isotropic turbulent flow. This comparison shows excellent prediction of our analysis at all Reynolds and Stokes numbers with a slight under-prediction when both Reynolds and Stokes numbers are high. The important role of Kolmogorov scale on particle clustering is confirmed by our analysis. [Preview Abstract] |
Monday, November 23, 2015 8:39AM - 8:52AM |
G3.00004: Fully resolved simulations of 2,000 fluidized particles Daniel Willen, Adam Sierakowski, Andrea Prosperetti Computational capabilities have matured sufficiently to render possible the dynamic simulation of thousands of resolved particles in fluid flows, generating an unprecendented amount of data. In this work we present a simulation of 2,000 fluidized particles generated with the Physalis method, and focus on probing the data with tools from statistical physics. In particular, the study of particle triads and tetrads has been used to study the dispersion of passive scalars in turbulence. Knowledge of the average shape and size of these structures over time provides insight into particle diffusion and the persistence of clusters. [Preview Abstract] |
Monday, November 23, 2015 8:52AM - 9:05AM |
G3.00005: Numerical Investigation of the Preferential Concentration Instability of Particle Laden Homogeneous Shear Mohamed Kasbaoui, Donald Koch, Olivier Desjardins In a previous study (Kasbaoui et al, J. Fluid Mech. 2015), particle laden homogeneous shear was shown to be subject to an algebraic instability. Initially randomly distributed particles are entrained by wave-like perturbations in the fluid velocity and segregate in a similar wave-like pattern while they sediment under gravity. The preferential concentration mechanism, which is the tendency of particles to exit vortical regions and gather in straining regions, causes the two waves to amplify each other resulting in an algebraic instability. By means of simulations, we compare the perturbations growth to the one yielded by the theory in the limit of small Stokes number particles. The simulations are conducted with an Eulerian model of the particles as well as a Lagrangian model. The two are compared. A secondary Rayleigh-Taylor instability caused by the periodic stacking of heavy layers of concentrated particles on top of depleted lighter layers is analyzed. [Preview Abstract] |
Monday, November 23, 2015 9:05AM - 9:18AM |
G3.00006: Particle dispersion in an inhomogeneous turbulent flow Peter Huck, Romain Volk An experimental study of the von K{\'a}rm{\'a}n swirling flow in counter rotation at $R_{\lambda} = 200$ is presented and investigates the interaction of tracer particles with a spatially inhomogeneous environment. Using ombroscopic PTV, eulerian conditioning of lagrangian trajectories permits visualization of the well known stagnation point topology and allows the calculation of lagrangian statistics which reveal the small scale anisotropy characteristic of this flow. Neighboring regions dominated by rotation or shear are also presented as anisotropy varies with respect to the predominate mean flow. The length of particle tracks needed to estimate velocity correlations is considerable and renders the rms velocity non stationary as particles explore their inhomogeneous surroundings. We conclude on the role of inhomogeneity and non stationarity in the concomitant process of particle dispersion. [Preview Abstract] |
Monday, November 23, 2015 9:18AM - 9:31AM |
G3.00007: Laboratory measurement of non-spherical particle rotation in turbulence: analysis in lab and local reference frames Ankur Bordoloi, Evan Variano Small axisymmetric particles are known to show shape dependence in their rotational kinematics in homogenous isotropic turbulence. For example, Byron et al. (2015) demonstrated that, rod-shaped particles rotate very differently compared to disc-shaped ones. This motivates us to extend this understanding to finite-sized particles ($\sim$ Taylor microscale) by examining their rotation and alignment. We have overcome the experimental challenges that have heretofore prevented the simultaneous measurement of orientation and rotation of large particles. We present this method and report results for large cylinders from 2D3C stereoscopic PIV data. These preliminary results show that there is equipartition of particle enstrophy into spinning about each of the particle's local axes. In other words, there is no preference for rotation about a particle's symmetry axis. Time permitting, effects of size on rotation for Taylor microscale particles will also be discussed. [Preview Abstract] |
Monday, November 23, 2015 9:31AM - 9:44AM |
G3.00008: Visualization of Air Particle Dynamics in an Engine Inertial Particle Separator Jason Wolf, Wei Zhang Unmanned Aerial Vehicles (UAVs) are regularly deployed around the world in support of military, civilian and humanitarian efforts. Due to their unique mission profiles, these advanced UAVs utilize various internal combustion engines, which consume large quantities of air. Operating these UAVs in areas with high concentrations of sand and dust can be hazardous to the engines, especially during takeoff and landing. In such events, engine intake filters quickly become saturated and clogged with dust particles, causing a substantial decrease in the UAVs' engine performance and service life. Development of an Engine Air Particle Separator (EAPS) with high particle separation efficiency is necessary for maintaining satisfactory performance of the UAVs. Inertial Particle Separators (IPS) have been one common effective method but they experience complex internal particle-laden flows that are challenging to understand and model. This research employs an IPS test rig to simulate dust particle separation under different flow conditions. Soda lime glass spheres with a mean diameter of 35-45 microns are used in experiments as a surrogate for airborne particulates encountered during flight. We will present measurements of turbulent flow and particle dynamics using flow visualization techniques to understand the multiphase fluid dynamics in the IPS device. This knowledge can contribute to design better performing IPS systems for UAVs. [Preview Abstract] |
Monday, November 23, 2015 9:44AM - 9:57AM |
G3.00009: Phoresis-induced clustering of particles in turbulence Lukas Schmidt, Itzhak Fouxon, Dominik Krug, Maarten van Reeuwijk, Markus Holzner We demonstrate phoresis-induced clustering of non-inertial particles in turbulent flows. Phoretic mechanisms such as thermophoresis, chemotaxis or diffusiophroesis are known to create a particle drift with respect to the fluid. Theory, based on the framework of weakly compressible flow, predicts that particles in turbulence streaked by salinity gradients experience a diffusiophoretic drift and will thus form particle cluster. An inclined gravity current setup is used to analyse clustering due to the diffusiophoretic effect in turbulent flow experimentally. Simultaneous 3D particle tracking velocimetry and laser induced fluorescent measurements provide the full Lagrangian velocity field and the local salt concentration in the observed 3D domain. Two independent methods show consistent evidence of the theoretically predicted particle clustering in turbulence. This clustering mechanism can provide the key to the understanding of spontaneous clustering phenomena such as the formation of marine snow in the ocean. [Preview Abstract] |
Monday, November 23, 2015 9:57AM - 10:10AM |
G3.00010: Clustering and relative velocity of heavy particles under gravitational settling in isotropic turbulent flows Guodong Jin, Guo-Wei He Clustering and intermittency in radial relative velocity (RRV) of heavy particles of same size settling in turbulent flows can be remarkably changed due to gravity. Clustering is monotonically reduced at Stokes number less than 1 under gravity due to the disability of the centrifugal mechanism, however it is non-monotonically enhanced at Stokes number greater than 1 due to the multiplicative amplification in the case that the proposed effective Kubo number is less than 1. Although gravity causes monotonical reduction in the rms of RRV of particles at a given Stokes number with decreasing Froude number, the variation tendency in the tails of standardized PDF of RRV versus Froude number is obviously different: the tails become narrower at a small Stokes number, while they become broader at a large Stokes number. The mechanism of this variation stems from the compromise between the following two competing factors. The mitigation of correlation of particle positions and the regions of high strain rate which are more intermittent reduces the intermittency in RRV at small Stokes numbers, while the significant reduction in the backward-in-time relative separations will make particle pairs see small-scale structures, leading to a higher intermittency in RRV at large Stokes numbers. [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. |
© 2023 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
1 Research Road, Ridge, NY 11961-2701
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700