Direct numerical simulations of particle-laden turbulent channel flows at friction Reynolds number $Re_\tau$ from $600$ to $2000$ have been performed to examine the near-wall particle streaks. Different from the well-observed small-scale particle streaks in near-wall turbulence of low $Re_\tau$ , the present results show large-scale particle streaks through the computational domain formed for relatively high-inertia particles at high $Re_\tau$ . Transferred by large-scale sweep and ejection events ( $Q^-$ ), these high-inertia particles preferentially accumulate in near-wall regions beneath the large-scale low-speed flow streaks observed in the logarithmic region. The corresponding Stokes numbers are associated with the lifetime of large-scale $Q^-$ structures, which increases as the Reynolds number grows. The small-scale particle streaks with a typical Stokes number $St_\nu \approx 30$ are mainly driven by the $Q^-$ structures in the buffer layer, whose lifetime is approximately $30$ in viscous time unit. Therefore, we propose a new structure-based Stokes number normalized by the lifetime of $Q^-$ structures of different scales. The relevant flow scales that control the formation of the large-scale particle streaks are parameterized by the structure-based Stokes number. The small-scale (large-scale) particle streaks are most prominent when the buffer-layer (large-scale) structure-based Stokes number approaches unity. The present findings reveal that formation of near-wall particle streaks is governed by the $Q^-$ structures of different scales, and the particles with different inertia respond efficiently to the $Q^-$ structures of corresponding scales with respect to the particle translational motion.
Preferential orientations of inertialess non-spherical particles are examined through three qualitatively different stages of a time-evolving Taylor-Green vortex flow. Despite an unexpected decorrelation between the vorticity vector and the direction of Lagrangian stretching, experienced by material fluid elements over a substantial time interval, prolate spheroids aligned with the Lagrangian stretching direction, whereas oblate spheroids aligned with the Lagrangian compression direction. We therefore infer that spheroidal tracers orient themselves relative to the Lagrangian history of the velocity gradients, defined by the left Cauchy-Green deformation tensor, rather than with the fluid vorticity vector. This preferential alignment persists all throughout the statistically unsteady flow field, and even in the inviscid and non-turbulent early stage of the time-dependent vortex flow. This explains the observed preferential spinning of rods and tumbling of disks, similarly as in homogeneous isotropic turbulence, even at the early stage when the flow is anisotropic and laminar. These preferred modes of particle rotation prevail all through the evolving flow, despite a surprisingly long time interval, during which the fluid vorticity decorrelates from the direction of Lagrangian stretching.
The existence of a quiescent core (QC) in the center of turbulent channel flows was demonstrated in recent experimental and numerical studies. The QC-region, which is characterized by relatively uniform velocity magnitude and weak turbulence levels, occupies about 40% of the cross-section at Reynolds numbers Re τ ranging from 1000 to 4000. The influence of the QC region and its boundaries on transport and accumulation of inertial particles has never been investigated before.Here, we first demonstrate that a QC is unidentifiable at Re τ = 180, before an in-depth exploration of particle-laden turbulent channel flow at Re τ = 600 is performed. The inertial spheres exhibited a tendency to accumulate preferentially in high-speed regions within the QC, i.e. contrary to the well-known concentration in low-speed streaks in the near-wall region. The particle wall-normal distribution, quantified by means of Voronoï volumes and particle number concentrations, varied abruptly across the QC-boundary and vortical flow structures appeared as void areas due to the centrifugal mechanism. The QC-boundary, characterized by a localized strong shear layer, appeared as a barrier, across which transport of inertial particles is hindered. Nevertheless, the statistics conditioned in QC-frame show that the mean velocity of particles outside of the QC was towards the core, whereas particles within the QC tended to migrate towards the wall. Such upward and downward particle motions are driven by similar motions of fluid parcels. The present results show that the QC exerts a substantial influence on transport and accumulation of inertial particles, which is of practical relevance in high-Reynolds number channel flow.
Inertial spheroids, prolates and oblates, are studied in a decaying Taylor–Green vortex (TGV) flow, wherein the flow gradually evolves from laminar anisotropic large-scale structures to turbulence-like isotropic Kolmogorov-type vortices. Along with particle clustering and its mechanisms, preferential rotation and alignment of the spheroids with the local fluid vorticity are examined. Particle inertia is classified by a nominal Stokes number [Formula: see text] which to first-order aims to eliminate the shape effect. The clustering varies with time and peaks when the physically relevant flow and particle time scales are of the same order. Low inertial ([Formula: see text]) spheroids are subjected to the centrifuging mechanism, thereby residing in stronger strain-rate regions, while high inertial ([Formula: see text]) spheroids lag the flow evolution and modestly sample strain-rate regions. Contrary to the expectations, however, spheroids reside in high strain-rate regions when the particle and flow time scales are comparable due to the dynamic interactions between the particles and the evolving flow scales. Moderately inertial ([Formula: see text]) prolates preferentially spin and oblates tumble throughout the qualitatively different stages of the TGV flow. These preferential modes of rotation correlate with parallel and perpendicular alignments of prolate and oblate spheroids, respectively, with the local fluid vorticity. However, for high inertial spheroids preferential rotation and alignment are decorrelated due to a memory effect, i.e., inertial particles require longer time to adjust to the local fluid flow. This memory effect is not only due to high particle inertia, as in statistically steady turbulence, but also caused by the continuously evolving TGV flow scales.
Rod- and disk-like particles preferentially align parallel and perpendicular, respectively, to the fluid vorticity, both at the early as well as later stages of the unsteady Taylor–Green vortex (TGV) flow. The early stage of the flow is laminar and comprises anisotropic large-scale Taylor–Green structures, while the later stages resemble homogeneous isotropic turbulence with Kolmogorov-type small-scale structures. The reason for the orientational behavior of inertialess spheroids in the early stage of the TGV-flow has been sought by examining the alignments of spheroidal particles, not only with vorticity but also with Lagrangian stretching and compression directions of the fluid elements in our earlier paper [Jayaram et al., “Alignment and rotation of spheroids in unsteady vortex flow,” Phys. Fluids. 33, 033310 (2021)]. This article is a sequel to the above paper in which the spheroids' alignments are studied locally, in contrast to the volume-averaged statistics studied previously, to observe the influence of the local flow field on the spheroidal alignment. It has been observed through our studies that the alignments vary periodically in space and these variations can be associated with the large-scale periodicity of the flow field originating from the initial conditions of the TGV flow. Additionally, the intense vortex stretching in the early stages of the flow evolution is seen to be largely influencing the orientation of the spheroids.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.