Session R11: Particle-Laden Flows: General (5:00pm - 5:45pm CST)

Dynamics of spheres falling in quiescent flows

Despite the apparent simplicity of a single sphere falling in a quiescent flow, far enough from Stokes conditions, a rich dynamics has been found. Simulations have shown the existence of several particle trajectory regimes in a quiescent flow, whose onsets are determined by particle-fluid density ratio ($\rho_s/\rho_f$) and Galileo number. These regimes are mainly characterised by the presence or absence of oscillations and the trajectory angle with respect to the gravity.\\ Here, experiments to test the validity of the aforementioned simulations are presented. In particular, the dynamics of heavy spheres with particle-fluid density ratios between $(1.1, 10)$ and Galileo numbers between $(150, 400)$ has been studied. We have performed an exploration of the different trajectory regimes modifying the Galileo number by changing the viscosity and using particles with several densities. While the trajectories are measured using a 3D PTV system.   

Toward a Mean Velocity Scaling in Variable Property Particle-Laden Channel Flow

In a particle-based solar receiver, dense particles in a turbulent duct flow are radiatively heated, in turn heating the surrounding air. When the particle loading and radiation intensity are large enough, there is significant gas expansion, which sets up an accelerating particle-laden flow with substantial density and viscosity variation. Using a suite of heated, particle-laden channel flow simulations, we characterize the turbulent mass, momentum, and energy transport associated with both the particle and fluid phases, as well as the fluctuating particle number density and transmitted radiation fields. In this complex flow, we find that the Van-Driest transformation is not able to collapse the mean velocity profiles, so we examine alternate scalings. The suite of channel flow simulations were conducted at $Re_{\tau}\approx235$, using Lagrangian point particles with moderate Stokes number ($St^+\approx7.5$) and particle mass loadings ranging from $10\%-200\%$. Radiation intensities were chosen so that the ratio of radiative heat flux to the sensible heat of the mixture range from $0.4 - 5.8$.   

Particle capture by drops in Turbulence

We examine the capture of sub-Kolmogorov inertial particles by deformable liquid droplets in dilute turbulent channel flow. To simulate this solid-liquid-liquid system, we exploit a Eulerian-Lagrangian methodology based on DNS of the Continuity, Navier-Stokes and Cahn-Hilliard equations and on the solution of the Lagrangian equation of particle motion. The carrier flow and the droplets have same density and viscosity. To model the particle-interface interaction, we consider a capillary force based on the liquid-liquid surface tension and the distance from the particle center to the nearest point on the fluid interface: This force is exerted only in close proximity of the (diffuse) interface and drives particle capture. Simulations with and without inter-particle collisions were performed to examine particle capture and subsequent accumulation at the interface, which are found to depend on the combined action of turbulence and particle inertia. Capture and trapping is higher for lower inertia particles, and accumulation is found to occur in the high-positive-curvature regions of the interface. This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Sklodowska-Curie grant agreement No 813948.   

Inertial particle dynamics with history effects at fixed computational cost

The unsteady dynamics of a small, spherical particle in a non-uniform flow is described by the Maxey-Riley (MR) equation. The Basset-Boussinesq term in the equation, which is an integral of the particle’s history weighted by time-dependent kernel, is known to impact transient dynamics but is often neglected or approximated due to its computational cost which grows linearly with the time of simulation. We address this by using an evolution equation for the history term due to Prasath et al.(2019). They showed that the MR equation with Basset-Boussinesq term can be posed as a Robin boundary condition to the 1D heat equation. This formulation and the closed-form expressions it affords, leads to a new time-iterative scheme to evaluate the full MR equation without approximation for generic nonlinear fluid flows. This method has a fixed computational cost per time-step and gives spectral accuracy, thus making it amenable to long-duration and multi-particle runs. We further reduce the cost associated with the iterative solver by employing only local velocity gradient information to compute the history effect. We present our numerical scheme to handle the evolution of a particle with history effects in 2D nonlinear flows.   

Effect of Center of Mass-Offset on Settling and Rising Spheres

We experimentally investigate the effect an offset center of mass has on rising and settling spheres in a still fluid. We find that the dynamics and kinematics of the particles are extremely sensitive to offsets as small as 1$\%$ of the particle radius. We uncover that the governing parameter for the particle behaviour is the particle Froude number, which is defined as the ratio between a rotational timescale of the particle and the characteristic time of the vortex shedding. It is found that the frequency and the amplitude of particle oscillations, as well as the particle drag coefficient vary strongly as a function of $Fr$. These effects are strongest for $0.08 \le Fr \le 0.14$, where resonance occurs between the rotational and vortex shedding frequencies. Furthermore, our results indicate that the particle drag does not correlate well with the amplitude of path oscillations but more sensitively depends on the amount of particle rotation. In rising particles, such rotations are enhanced by the Magnus lift force resulting in an increase in drag. Contrary to this, for settling particles the Magnus lift force counteracts the particle rotation, such that no drag increase is observed in this case.   

Targeted particle delivery via reconnecting vortex rings

We propose and characterise a method for targeted delivery of finite-sized particles within a streamwise evolving channel flow. By randomly seeding particles at specific Stokes numbers within the core of vortex rings, the solid-phase is transported in the streamwise direction of the channel flow via self-advection of the vortex ring. The streamwise particle advection can then be halted and transferred into the wall-normal direction through anti-parallel vortex reconnection. To this end, a pair of nearly co-aligned vortex rings are setup to reconnect at a streamwise distance, thus enabling a transfer of the solid particles to the side walls of the channel. This talk will propose a systematic approach towards targeted particle delivery and quantify the mass transport, through high-fidelity direct numerical simulations, to the side walls of a laminar channel flow. This approach can be extended for use in drug delivery, surface treatment of internal flow passages, and food processing industry.   

Maximum drag reduction asymptote for particulate pipe flows

Adding polymers or particles to a flow can alter the drag experienced by it. For instance, polymers in a flow reduce drag but their ability is bounded by a limit, the maximum drag reduction asymptote. However, the effect of particles on drag is ambiguous, with studies reporting contradicting observations; even in cases where particles are reported to reduce drag no asymptotic limit is know. The ambiguity arises because in addition to particle concentration, particle shape, size and density affect the drag. Hence, various particles behave differently in distant ranges of Re and concentration. In the present study, we experimented in a pipe flow setup with neutrally buoyant spherical and elongated particles, covering wide ranges of both Re and concentration. Based on friction factor, we found that spherical particles do not show drag reduction at any Re while the elongated particles do within an specific interval of Re. This interval strongly depends on particle concentration and relative size of pipe to particle, and within this interval the friction factor reaches a minimum value. These drag reduction maxima fall onto a distinct curve which can be considered the maximum drag reduction asymptote for a given particle shape, irrespective of the pipe diameter or concentration.   

Methods for improving the particle sizing resolution of inertial impactors using ring shaped deposits

Inertial impactors are devices used to measure the size distribution of particles in a flow. These instruments are robust and widely used, however the number of bins in the resulting size distributions are limited by the number of impactors in an impactor cascade. Recent research shows that under proper conditions, the normally disc shaped deposition pattern takes the form of a ring whose diameter can be correlated to the particle diameter. This result opens the door to much higher resolution particle size distributions for an impactor cascade, or even for a single impactor. Herein, experiments are presented revealing these ring-shaped deposition patterns on glass substrates for a range of particle sizes, further illustrating the effect. The ring diameter is correlated to the particle diameter and Stokes number. Additionally, computer simulations of particle trajectories are presented, revealing the underlying mechanism of this effect.   

Dilute suspensions over and through porous structures; the impact of permeability and porous thickness

Suspension flows at various concentrations have been studied significantly within impermeable/smooth channels. Likewise, porous media has been studied in various engineering applications including carbon nanotube forests applications. However, there are few studies in terms of the coupling between these two flows. This study examines a pressure-driven, dilute suspension flow in a channel where the bottom wall replaced by porous media. The properties of the porous media were manipulated to see how they change the velocity profiles and parameters at the suspension-porous interface. A dilute, non-Brownian, neutrally-buoyant suspension of rigid spherical particles with 3{\%} bulk volume fraction was passing over rigid porous media consisting of cylindrical rods perpendicular to the flow direction. It was observed that as the permeability of porous media increases the location of the maximum velocity within the free-flow region shifts towards the interface. Properties at the interface such as, the dimensionless slip parameter, slip length, slip coefficient, and penetration depth are all greatly affected by the porous media thickness and permeability.   

Novel Indirect Particle Temperature Measurement Methodology Using IR and Visible-light Cameras

The particle temperature measurements in a gravity-driven flows present a unique challenge due to its transient nature and flow's stochastic. While attempts to estimate the bulk particle temperature have been conducted using contact and non-contact methods, a definitive and practical solution is yet to be found. This work focuses on a novel non-contact method using a high-speed IR camera and a visible-light camera (Nikon D3500) to accomplish this indirect particle temperature measurement. The thermograms and image sets collected by the cameras allow for the measurement of the apparent particle temperature and the opacity of a particle plume. An in-house post-processing code based on Planck's theory allows to calculate the true particle temperature from the apparent temperature obtained from the thermograms. The particle temperature data are compared with the empirical model of the bulk particle temperature yielding agreement with 95{\%} confidence (2$\sigma )$.   

Electron Temperature Measurement using Collisional Radiative Model in PFRC-II

Replace this text with your abstract body.