Session A10: Particle-Turbulence Interactions I

Direct numerical simulation of hydrosol dynamics in turbulent channel flow

Deposition of Asphaltene particles in crude oil during production/process poses a serious challenge in petroleum industry and has gone unaddressed yet. More specifically, dispersion and deposition of hydrosol (solid particles in a liquid such as asphaltene particles in crude oil) have not been studied extensively, and may have different dynamics compared to widely studied aerosol (solid particles in a gas) dynamics due to different properties of the carrier (fluid) phase such as density. Here we aim to take a foundational step to unveil the major underlying mechanisms of asphaltene particle (hydrosol) deposition process using high fidelity data obtained from Point-Particle Direct Numerical Simulations (PP-DNS). In particular, dispersion and deposition dynamics (e.g., deposition velocity) of hydrosol will be assessed and compared to aerosol when small particles are considered.   

Physics-informed neural networks for synthesizing preferential concentration of particles in turbulent flows

Cluster and void formation are key processes in the dynamics of particle-laden turbulence. Here we consider direct numerical simulation (DNS) of inertial point particles in homogeneous isotropic turbulence at high Reynolds numbers as the database. We propose different data driven and physics-informed machine learning techniques for synthesizing preferential concentration fields. For training, we are using the enstrophy fields computed by DNS and compare the accuracy using the enstrophy either in the physical space or a sparse wavelet space representation of vorticity. We propose to use and compare autoencoder, U-net, and generative adversarial network (GAN) approaches. We assess the statistical properties of the generated fields, and we find that the best results, showing clusters and voids, are obtained with GANs. This yields interesting perspectives for reducing the computational cost of expensive DNS computations by avoiding the tracking of billions of particles. We also explore the inverse problem of synthesizing the enstrophy fields using the particle density distribution, at different Stokes numbers, as the input. Our study thus provides perspectives to use neural networks to predict the enstrophy field using particle data in experimental PIV measurements.   

Coupling of anisotropic fiber tracking with instantaneous volumetric flow field in turbulent channel flows.

Time-resolved and instantaneous volumetric measurements are coupled with Lagragian tracking of anisotropic curved fibers in the TU Wien Turbulent Water Channel. The experiments are conducted at friction Reynolds number ranging from 180 to 720 in the viscous wall region and in the channel center. The length of the fiber (1.2mm) and the spatial resolution of the velocity vector field (0.5mm) are comparable to the Kolmogorov length scales (0.3-0.8 mm). Full rotation rate tensor of small slender fibers is simultaneously captured with instantaneous velocity gradient tensor of the flow, which allows the determination of statistical and instantaneous velocity and rotational slip between the fiber and the flow structures (laden).

The measurement technique adopted in this work represents an improvement in respect to previous experimental investigations (Alipour et al., 2021 and 2022). As results, the different behaviors of the fibers in terms of transnational and rotational motion in different regions of the channel are correlated to the homogeneous turbulent structures in the center of the channel and to the coherent structures within the viscous wall region, i.e., hairpin packets and transverse and longitudinal vortices.   

Effect of low clay concentration on nearly isotropic turbulence

An experimental investigation was conducted to investigate and quantify the effect of low clay concentrations on nearly isotropic turbulence created in a customized cubic mixing box. A novel approach using Laponite RD, a synthetic clay, produced a nearly transparent water-clay mixture suspension. This clay-laden flow allowed full optical access for high-speed particle image velocimetry (PIV) to characterize the spatio-temporal dynamics around the center of the mixing box. We investigated mass concentrations of C(%) = 0, 1, 1.5, and 2, which minimized non-Newtonian behavior. Clay particles significantly modulated the turbulence structure even at low concentrations of laponite with a monotonic reduction of turbulence kinetic energy, mean dissipation rate, and an increase of the Taylor microscale, λ, as clay concentration increased. Velocity spectra demonstrated a clear dependence on clay concentration and the Taylor microscale Reynolds number scaling correction, p, of the power law in the inertial subrange. The Taylor microscale estimation by p-factor agrees well with those obtained from structure function and direct PIV measurements. The energy budget suggests that the reduction in TKE and increase in λ are likely associated with the presence of clay particle aggregation.   

Transport of Inertial Droplets in a Turbulent Boundary Layer

The transport of highly inertial droplets in the presence of a turbulent boundary layer is studied experimentally inside a wind tunnel. Water droplets are injected into the boundary layer at free-stream velocities ranging from U_∞ = 2.4 to 6.6 m/s with a boundary layer thickness ranging from δ = 11.4 to 12.5 cm. The size, speed, and acceleration of droplets are measured using a high-speed cinematic in-line holographic system positioned 162 cm downstream from the nozzle and at many wall-normal streamwise locations spanning the full thickness of the boundary layer. Water droplets with diameters ranging from 30 to 1000 μm are measured with corresponding Stokes numbers ranging from St_η = 0.11 to 160. Wall-normal droplet concentration and velocity profiles are reported. Droplet statistics like velocity and acceleration are discussed as a function of droplet size and wall-normal location. The importance of turbulence on the motion of highly inertial droplets is discussed and compared to similar statistics obtained from inertial droplets produced by breaking wind-forced waves.   

On the suspended microplastic dispersion in turbulent shallow flows under progressive gravity waves

When wind waves approach the coast over a relatively shallow water, the wave-induced orbital motion promotes the development of a turbulent boundary layer at the bottom, enhancing the transport and dispersion of suspended microplastics. Furthermore, the strong wave-bed interaction results in a net non-periodic shoreward Longuett-Higgins current close to the bottom superimposed to the wave orbital motion [1]. Via direct numerical simulation, we analyse the effect of the wave orbital motion, the steady current and the near-wall turbulence on the dispersion and vertical distribution of microplastics both inside and outside the bottom boundary layer. For performing the simulations, we utilise a boundary-fitted hybrid pseudo-spectral/finite-difference code which implicitly solves the Navier-Stokes equations together with the dynamic and kinematic boundary conditions at the free-surface. The motion of the suspended microplastics is treated by a Lagrangian particle tracking algorithm in the framework of the point-particle approach. The aim of this work is to understand the physical mechanisms behind the microplastic transport processes under shallow progressive waves. This has important implication in parameterising the mechanistic behaviour of suspended microplastics and assessing existing predictive microplastic transport models.

[1] Longuet-Higgins, M. S. (1953). Mass transport in water waves. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences, 245(903), 535-581.   

Modelling precipitation-driven flows in planetary interiors: effects of particle inertia and dissolution

As an analog of settling snow flakes in planetary interiors which may drive large scale motions in liquid metal, we instantaneously release clouds of glass spheres in a water tank, which quickly evolve as turbulent thermals. Synchronous cameras analyse the two-way coupling between the settling spheres and the fluid : one camera performs a Lagrangian tracking of particles while the other one analyses turbulence with PIV or LIF. The size of particles is systematically varied in the range 5µm-1mm. We notably show that due to particle-turbulence interactions, particle clouds entrain more than salt water clouds of identical mass excess, with an optimum of entrainment for a finite inertia, as confirmed by our 3D DNS of the same clouds using Basilisk.

Subsequently, glass spheres are replaced with grains of custom-made sugar to model melting of the snow flakes by dissolution in water. Grains of controlled size are continuously sieved above water with a chosen mass rate. As grains fall and dissolve, they impart momentum and buoyancy to the fluid. We analyse the growth of the resulting mixing layer, which eventually collapses due to Rayleigh-Taylor instabilities, producing a macroscopic plume which carries grains to large depths and nourishes an inhomogeneous compositional convection.   

Surfacing of gyrotactic swimmers in stratified/unstratified free-surface turbulence

The surfacing of gyrotactic swimmers in free-surface turbulence mimics the dynamic behavior of several phytoplankton species in water bodies in the absence of large surface waves and ripples. We examine the surfacing dynamics via direct numerical simulation of open-channel turbulence at varying shear Reynolds number (from 170 to 1018 based on the channel height) coupled with Lagrangian tracking of inertialess swimmers with stability number covering 2 orders of magnitude, which orient themselves according to Jeffery equation. We consider both spherical and elongated swimmers, the latter being modeled as prolate ellipsoids with aspect ratio up to 10. We show that the vertical migration of the swimmers, which are initially distributed randomly throughout the flow, depends on both their re-orientation ability and shape. Once at the surface, swimmers tend to stay there and organize themselves into filamentary clusters, which are in direct correlation with the topology of surface turbulence. The surfacing process is found to depend strongly on the level of thermal stratification within the water body, which we examine at varying Richardson number (from 0, corresponding to the unstartified case, to 510, corresponding to a stratfied flow characterized by the occurrence of a thermocline below the free-surface). Our results indicate that free-surface turbulence and thermal stratification have a strong influence on the swimmers' spatial distribution and, in turn, on the carbon-dioxide exchange cycle across the water/atmosphere interface.   

Effect of particle size on characteristics of transitional particle-laden pipe flow

Neutrally buoyant particles in laminar pipe flow exhibit a phenomenon called the tubular pinch effect, where they accumulate at a certain radius near the wall. Particles also promote or hinder the transition to turbulence. The transitional regime is marked by intermittent turbulent structures called puffs. These phenomena are characterized by the particle-to-pipe diameter ratio(D/d), the particle volume fraction (ϕ), and Reynolds Number (Re). In the current study, we examine the interaction between polystyrene beads in a 20% glycerol-water solution (ρ=1046 kg-m^-3) and puffs using PTV. Even at small loadings (ϕ=0.2%), larger particles (D/d=43) lead to transition at lower Re than single-phase flow while smaller particles (D/d=86) delay the transition to higher Re. In the laminar regime, the particle concentration was found to peak near 0.88R. Puffs disrupt this accumulation, with the particles migrating radially inwards. The radial velocities of the particles are significant over a distance of 10D near the trailing edge of the puff, and the disruption persists for more than 100D. We examine the radial trends of particle velocity, concentration, and recovery with respect to particle size and volume fraction.