Hostname: page-component-7b7567c8c-fvjsw Total loading time: 0 Render date: 2026-11-16T05:23:06.252Z Has data issue: false hasContentIssue false

Near-wall dynamics of a neutrally buoyant spherical particle in an axisymmetric stagnation point flow

Published online by Cambridge University Press:  06 April 2020

Qing Li
Affiliation:
Laboratoire de Génie Chimique, CNRS-Toulouse INP-UPS Université de Toulouse, 4 Allée Emile Monso, 31432Toulouse, FranceInstitut de Mécanique des Fluides de Toulouse, CNRS-Toulouse INP-UPS Université de Toulouse, Allée Camille Soula, 31400Toulouse, FranceFR FERMAT, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
Micheline Abbas*
Affiliation:
Laboratoire de Génie Chimique, CNRS-Toulouse INP-UPS Université de Toulouse, 4 Allée Emile Monso, 31432Toulouse, FranceFR FERMAT, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
Jeffrey F. Morris
Affiliation:
Benjamin Levich Institute and Department of Chemical Engineering, The City College of New York, New York, NY10031, USA
Eric Climent
Affiliation:
Institut de Mécanique des Fluides de Toulouse, CNRS-Toulouse INP-UPS Université de Toulouse, Allée Camille Soula, 31400Toulouse, FranceFR FERMAT, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
Jacques Magnaudet
Affiliation:
Institut de Mécanique des Fluides de Toulouse, CNRS-Toulouse INP-UPS Université de Toulouse, Allée Camille Soula, 31400Toulouse, FranceFR FERMAT, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
*
†Email address for correspondence: micheline.abbas@ensiacet.fr

Abstract

The motion of a neutrally buoyant spherical particle along the axis of an axisymmetric stagnation point flow at a rigid and smooth flat wall (Hiemenz–Homann flow) is investigated in the presence of low-to-moderate inertia effects. The particle dynamics is elucidated using numerical simulation. At distances large compared to the characteristic thickness of the boundary layer $\unicode[STIX]{x1D6FF}=(\unicode[STIX]{x1D708}/B)^{1/2}$, with $\unicode[STIX]{x1D708}$ the kinematic viscosity and $B$ the strain rate of the carrying flow, the particle decelerates as it approaches the wall, due to the ambient pressure increase toward the stagnation point. In this part of the path, its velocity is nearly identical to that of the local undisturbed fluid at the position of its centre. Relative motion between the particle and fluid increases as the wall–particle gap reduces, due to wall-induced hydrodynamic interaction forces. Two distinct evolutions of the net force on the particle are observed, depending on the relative particle size, $a/\unicode[STIX]{x1D6FF}\sim Re^{1/2}$, where $a$ is the particle radius and $Re=2Ba^{2}/\unicode[STIX]{x1D708}$ is the Reynolds number. For $a/\unicode[STIX]{x1D6FF}\lesssim 2$, the force decays monotonically to zero, while it undergoes a sharp rise before returning to zero for larger particles. In the latter case, the particle retains a sufficient velocity even for very small gap widths such that, under usual roughness levels, a rebounding collision would occur. The stress profiles at the particle surface are investigated to separate the various contributions to the hydrodynamic force. Theoretical predictions for near-wall viscous and inertial forces available in the creeping-flow and low-but-finite Reynolds-number limits, respectively, are used to pinpoint the origin of the dominant inertia effect that controls the particle dynamics when the particle gets very close to the wall.

Information

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

Figure 1.Sketch of a rigid spherical particle transported on the axis of an axisymmetric HH flow, some streamlines of which are displayed. (a) Vorticity iso-levels (vorticity is scaled by the flow strain rate, $B$); (b) pressure iso-levels (pressure differences with respect to the pressure at infinity, $P_{\infty }$, are scaled by the pressure difference at the stagnation point, $P_{0}-P_{\infty }$, determined in the single-phase HH flow). The outer edge of the boundary layer, where the radial fluid velocity component reaches 98 % of its free-stream value, stands a distance $3\unicode[STIX]{x1D6FF}$ from the wall.

Figure 1

Figure 2.Comparison between computational predictions (dashed lines) and the theoretical HH solution (solid coloured lines) for the undisturbed fluid velocity distribution. (a) Radial velocity profiles at various radial positions, namely $r/\unicode[STIX]{x1D6FF}=1.54$ (blue), 3.12 (magenta) and 7.86 (red); (b) wall-normal velocity profile at $r/\unicode[STIX]{x1D6FF}=1.54$. The grid is identical to that used in cases $c_{1}{-}d_{0}$ defined in table 1.

Figure 2

Table 1.Parameters used in the simulations. Second column: particle radius scaled by $\unicode[STIX]{x1D6FF}$; third and fourth columns: normalized minimum grid size in the axial and radial directions, respectively; fifth column: normalized time step.

Figure 3

Figure 3.Ambient inertial force $F_{z}^{\infty }$ acting on a fluid sphere with radius $a$ standing a distance $z_{c}$ from the wall, as a function of the dimensionless gap $\unicode[STIX]{x1D716}=h/a$, with $h=z_{c}-a$. Red, green, blue and black lines correspond to $a/\unicode[STIX]{x1D6FF}=0.8$, 1.6, 2.4 and 3.2, respectively.

Figure 4

Figure 4.Flow streamlines (in the wall reference frame) about two particles moving toward the wall, at wall-normal positions $z_{p}/a=1.5$ (a,b) and $z_{p}/a=1.01$ (c,d). (a,c) ‘Small’ particle with $a/\unicode[STIX]{x1D6FF}=0.8$; (b,d) ‘large’ particle with $a/\unicode[STIX]{x1D6FF}=3.2$. The dashed line indicates the distance $z/a=\unicode[STIX]{x1D6FF}/a$ from the wall.

Figure 5

Figure 5.Particle velocity and particle–fluid slip velocity as a function of the distance to the wall. (a) Particle velocity versus the normalized wall-normal position of the particle centre $z_{p}/\unicode[STIX]{x1D6FF}$; (b) slip velocity scaled by the fluid velocity at $z=a$ ($\unicode[STIX]{x1D716}=0$), as a function of the dimensionless gap. $a/\unicode[STIX]{x1D6FF}=3.2$ (black line), 2.4 (dotted blue line), 1.6 (dash-dotted green line), 0.8 (red line); blue solid line in (a): tracer-like particle; black dashed line in (b): $a/\unicode[STIX]{x1D6FF}=0.1$. The pink line in (b) corresponds to the theoretical prediction for the slip in the Stokes flow limit obtained with $a/\unicode[STIX]{x1D6FF}=0.01$ and $\unicode[STIX]{x1D716}\rightarrow 0$ (Rallabandi et al.2017).

Figure 6

Figure 6.Two characteristics of the differential motion between the particle and fluid for different particle sizes: (a) slip Reynolds number, $Re_{s}$; (b) relative acceleration, $(\text{D}U/\text{D}t)-(\text{d}V/\text{d}t)$, normalized by $a_{\unicode[STIX]{x1D6FF}}=\unicode[STIX]{x1D6FF}B^{2}$. $a/\unicode[STIX]{x1D6FF}=3.2$ (black line), 2.4 (dotted blue line), 1.6 (dash-dotted green line), 0.8 (red line).

Figure 7

Figure 7.Variation of the forces experienced by the particle with respect to the wall-normal position. (a) Total force, $F_{T}$; (b) hydrodynamic force, $F_{H}=F_{T}-F_{z}^{\infty }$. In (b), symbols indicate the value taken by $F_{H}$ when the particle is held fixed at the stagnation point (the colour code is identical to that used for the force profiles). The inset shows the variations of $F_{H}$ for the particle corresponding to $a/\unicode[STIX]{x1D6FF}=0.8$ very close to the wall. The line style refers to the particle size: $a/\unicode[STIX]{x1D6FF}=3.2$ (black line), 2.4 (dotted blue line), 1.6 (dash-dotted green line), 0.8 (red line).

Figure 8

Figure 8.Variation of the ‘hydrodynamic’ force $F_{H}$ with the particle Reynolds number at the location corresponding to $\unicode[STIX]{x1D716}=1$ for each particle in the range $0.1\leqslant a/\unicode[STIX]{x1D6FF}\leqslant 3.2$. Error bars correspond to the maximum deviation observed when the particle initial position $z_{p0}/\unicode[STIX]{x1D6FF}$ is varied from 12 to 20.

Figure 9

Figure 9.Effect of the near-wall numerical resolution on the total force $F_{T}$ acting on the largest particle ($a/\unicode[STIX]{x1D6FF}=3.2$) at small gaps. The line style refers to the grid in table 1: $d_{2}$ (black), $d_{1}$ (red solid), $d_{0}$ (red dashed). Additional under-resolved runs with a minimum cell size $\unicode[STIX]{x1D6E5}_{z}^{m}/\unicode[STIX]{x1D6FF}=0.08$ (blue) and 0.04 (green), both with $B\unicode[STIX]{x0394}t=1\times 10^{-4}$.

Figure 10

Figure 10.Angular distribution of the contribution $f(\unicode[STIX]{x1D703})$ to the hydrodynamic force at various gaps and for three particle sizes; $\unicode[STIX]{x1D703}$ is in degrees from the position of closest approach to the wall. (a) $a/\unicode[STIX]{x1D6FF}=3.2$, (b) 2.4 and (c) 0.8, with $\unicode[STIX]{x1D716}=0.005$ (green), 0.01 (black), 0.05 (blue), 0.10 (red) and 0.20 (magenta). Inset: $F_{H}$ computed from the right-hand side of (2.3) (circles) or defined as $F_{H}=\int _{0}^{\unicode[STIX]{x03C0}}f(\unicode[STIX]{x1D703})\,\text{d}\unicode[STIX]{x1D703}$ (squares).

Figure 11

Figure 11.Pressure profiles along the thin-gap region of the particle surface at $\unicode[STIX]{x1D716}=0.01$, as a function of the angle in degrees measured from $\unicode[STIX]{x1D703}=0$ on the axis of the motion. Solid lines: numerical predictions; dashed lines: thin-gap approximation (4.2), both for $a/\unicode[STIX]{x1D6FF}=3.2$ (black), 2.4 (blue), 1.6 (green) and 0.8 (red).

Figure 12

Figure 12.Force profiles in the thin-gap region for (a) $a/\unicode[STIX]{x1D6FF}=0.8$, and (b) $a/\unicode[STIX]{x1D6FF}=3.2$. Hydrodynamic force $F_{H}$ (green), lubrication force ${\mathcal{F}}_{L}$ (red dashed) and sum ${\mathcal{F}}_{0}=F_{{\mathcal{A}}}+F_{{\mathcal{B}}}+F_{{\mathcal{C}}}$ of the three forces involved in the creeping-flow prediction (4.1) (black solid). The blue lines correspond to individual contributions to (4.1): $F_{{\mathcal{A}}}$ (solid), $F_{{\mathcal{B}}}$ (dashed) and $F_{{\mathcal{C}}}$ (dotted). Inset in (a): zoom on the $F_{H}$, ${\mathcal{F}}_{L}$ and ${\mathcal{F}}_{0}$ profiles.

Figure 13

Figure 13.Validation of the numerical approach in the case of a particle settling toward a plane wall (the inset zooms on the region $\unicode[STIX]{x1D716}\leqslant 2\times 10^{-2}$). Symbols: experimental data from Mongruel et al. (2010) for $St=9.24$ (blue), 6.90 (red), 3.90 (green) and 1.72 (magenta); solid lines: numerical results. The Stokes velocity $V_{St}$ and Stokes number $St$ are defined as $V_{St}=2(\unicode[STIX]{x1D70C}_{p}/\unicode[STIX]{x1D70C}-1)ga^{2}/9\unicode[STIX]{x1D708}$ and $St=2(\unicode[STIX]{x1D70C}_{p}/\unicode[STIX]{x1D70C})aV_{St}/9\unicode[STIX]{x1D708}$, respectively.

Figure 14

Figure 14.Schematic of the axisymmetric biaxial straining flow past a spherical particle.

Figure 15

Figure 15.Velocity disturbance induced by a sphere held at the hyperbolic point of a biaxial straining flow: (a) axial velocity along the $z$-axis, (b) radial velocity within the $(x,y)$-diametrical plane. Black line: theoretical solution, red line: simulation results.

Figure 16

Figure 16.Same as figure 15 within the diagonal plane inclined by $+45^{\circ }$ with respect to the $(x,y)$-plane. Black line: theoretical solution, blue line: simulation results.