Hostname: page-component-7b7567c8c-tnkh4 Total loading time: 0 Render date: 2026-11-17T01:47:13.923Z Has data issue: false hasContentIssue false

Falling clouds of particles in vortical flows

Published online by Cambridge University Press:  10 December 2020

Benjamin Marchetti
Affiliation:
Aix Marseille Université, CNRS, IUSTI, 13453 Marseille, France
Laurence Bergougnoux*
Affiliation:
Aix Marseille Université, CNRS, IUSTI, 13453 Marseille, France
Elisabeth Guazzelli
Affiliation:
Université de Paris, CNRS, Matière et Systèmes Complexes (MSC) UMR 7057, 75205 Paris, France
*
†Email address for correspondence: laurence.bergougnoux@univ-amu.fr

Abstract

The coupling between particle–particle and particle–fluid interactions is examined by studying the sedimentation of clouds of spheres in a model cellular flow at a small but finite Reynolds number. The model flow consists of counter-rotating vortices and is aimed at capturing key features of the vortical effects on particles. The dynamics of clouds settling in this vortical flow is investigated through a comparison between experiments and point-particle simulations.

Information

Type
JFM Papers
Copyright
© The Author(s), 2020. Published by Cambridge University Press
Figure 0

Figure 1.Experimental apparatus. The transparent cell is filled with an aqueous mixture of citric acid (UCON$^{\scriptsize {\circledR}}$ oil can be added to increase the viscosity). The magnetic field is generated by a checkerboard of magnets placed behind the back wall of the cell. The electric current is created by two electrodes placed on the opposite small sides of the cell. The flow generated by this set-up is a periodic flow of counter-rotating vortices that is depicted in figure 2. Clouds are dropped from the top of the cell using a custom made syringe and settle through the cellular flow. Their trajectories are imaged by a set of synchronised cameras located below and in front of the cell.

Figure 1

Figure 2.Vortical flow: velocity-vector map of the flow field in the vertical plane ($zy$) for different horizontal positions ($x$) and in the horizontal plane ($xy$) for different vertical positions ($z$) in cases (a) $A1a$ ($Re_k=0.7$) and (b) $C2b$ ($Re_k=13.6$).

Figure 2

Table 1.Particle characteristics: mean sphere radius $a$ (the error bar corresponds to the standard deviation) and density $\rho _p$. Batches $A$ and $B$ are made of polystyrene particles while batch $C$ consists of polymethyl methacrylate (PMMA) particles.

Figure 3

Table 2.Fluid characteristics: viscosity $\mu$ and density $\rho _f$.

Figure 4

Table 3.Dimensionless numbers for the different combinations of particles (labelled $A, B$ and $C$), fluids (labelled 1 and 2) and current intensities (labelled $a$ and $b$) used in the experiments.

Figure 5

Figure 3.Typical three-dimensional trajectories of centres of mass of clouds. The colour coding corresponds to the distance to the back wall, and the trajectories projected in the horizontal plane are shown at the bottom. The trajectories which do not possess a large $x$-extension are selected.

Figure 6

Figure 4.Typical (a) experimental and (b) numerical chronophotographs of a cloud for case $A1b$. The time between successive photos is kept constant (2 s in the experiments and a corresponding interval in the simulation) in order to indicate the difference in velocities along the trajectory. The flow field measured by PIV is indicated by red arrows. The complete dynamics for the experimental case and for the Stokeslet simulation are shown in supplementary movie 1 available at https://doi.org/10.1017/jfm.2020.883.

Figure 7

Figure 5.Same as figure 4 but for case $B1b$. The complete dynamics for the experimental case and for the Stokeslet simulation are shown in supplementary movie 2.

Figure 8

Figure 6.Same as figure 4 but for case $C2b$. The numerical simulation uses an Oseenlet (instead of Stokeslet) approximation. The complete dynamics for the experimental case and for the Oseenlet simulation are shown in supplementary movie 3.

Figure 9

Figure 7.Typical experimental ($\circ$) and numerical (solid lines) trajectories in the Stokes case $B1a$: time evolution of the (a) vertical, $Z$, and (b) horizontal, $Y$, coordinates of the cloud centre of mass, (c) its vertical velocity, $U_z$, and (d) the cloud projected surface in the horizontal plane normalised by its initial value, $\varSigma /\varSigma _0$. In (a), the black triangle indicates the slope $U_c = N_0 U_S 6a/5R_c$ and the inset shows the initial positions of the cloud within experimental accuracy $\varDelta$ (with $k \varDelta =7\times 10^{-2}$). The horizontal dotted lines correspond (i) to the initial location $kY=0$ in (b), (ii) to the value $U_c/U_0$ in (c) and (iii) to the initial projected surface $\varSigma /\varSigma _0=1$.

Figure 10

Figure 8.Same as figure 7 but for the finite-inertia case $C2b$. The $\star$ symbols indicate break-up events.

Figure 11

Figure 9.Snapshots of the settling numerical cloud for cases (a) $B1a$ ($N_0=2500$) and (b) $C2b$ ($N_0=500$). The original flow field measured by PIV is indicated by black arrows while the perturbed flow field is represented by blue streamlines. The corresponding movies are available in the supplementary material (movie 4 for the Stokeslet simulation and movie 5 for the Oseenlet simulation).

Figure 12

Figure 10.Percentage $(N_0-N)/N_0$ of particles that have leaked away from the cloud as a function of dimensionless time $tU_c/R_c$: in (a) quiescent fluids and (b) vortical flows for cases $A1a$, $B1b$, and $C2b$; in vortical flows for different $N_0$ for cases (c) $B1b$ and (d) $C2b$; in vortical flows for two different $Re_k$ for cases (e) $B1b$ and (f) $C2b$. The insets of (c,d) show the same leakage $N_0-N$ but normalised by $N_0^{1/3}$ instead of $N_0$.

Figure 13

Table 4.Dimensionless break-up times, $t_b U_c/R_c$, in the quiescent and vortical cases ($C2a$ and $C2b$). The error bars correspond to the standard deviations between the different experimental or numerical runs.

Figure 14

Figure 11.Flow field computed at successive times in the vertical plane in the cloud reference frame for a typical Oseenlet simulation in case $C2b$. High (low) velocity is indicated in white (dark). The corresponding movie 6 is available in the supplementary material.

Figure 15

Figure 12.Snapshots of numerical clouds (using the Oseenlet approximation) corresponding to case $C2a$ ($P=0.03, W=0.108,\ell /R_c=2$) settling in Taylor–Green vortices at five different times $\hat {t}=t/(kU_0)^{-1}=0$, 8, 16, 24, 32 (each time $\hat {t}$ corresponds to a different colour) with (a) $Q=0.5$ and $N_0=500$, (c) $Q=2$ and $N_0=500$, and (e) $Q=2$ and $N_0=2000$. Conditions in (b), (d), (f) are the same as in (a), (c), (e), respectively, but hydrodynamic interactions between particles have been switched off. The corresponding movie 7 is available in the supplementary material.

Marchetti et al. supplementary movie 1

See pdf file for movie caption

Download Marchetti et al. supplementary movie 1(Video)
Video 792.5 KB

Marchetti et al. supplementary movie 2

See pdf for movie caption

Download Marchetti et al. supplementary movie 2(Video)
Video 1.3 MB

Marchetti et al. supplementary movie 3

See pdf file for movie caption

Download Marchetti et al. supplementary movie 3(Video)
Video 831 KB

Marchetti et al. supplementary movie 4

See pdf file for movie caption

Download Marchetti et al. supplementary movie 4(Video)
Video 5.4 MB

Marchetti et al. supplementary movie 5

See pdf file for movie caption

Download Marchetti et al. supplementary movie 5(Video)
Video 2.8 MB

Marchetti et al. supplementary movie 6

See pdf file for movie caption

Download Marchetti et al. supplementary movie 6(Video)
Video 3.6 MB

Marchetti et al. supplementary movie 7

See pdf file for movie caption

Download Marchetti et al. supplementary movie 7(Video)
Video 585 KB
Supplementary material: PDF

Marchetti et al. supplementary material

Captions for movies 1-7

Download Marchetti et al. supplementary material(PDF)
PDF 49.7 KB