• Home
• Program
• Contact

16:15   Transport and Mixing 2 16:15 15 mins #137 BUOYANCY REGULATION OF NON-MOTILE PHYTOPLANKTON IN A TURBULENT FLOW Matteo Borgnino Abstract: Phytoplankton cells have developed non-trivial strategies to actively respond to environmental signals such as incident light, nutrient concentration or mechanical stresses. All these stimuli affect microorganisms dynamics and distribution over different scales. In recent years, significant efforts have been devoted to understand the mechanisms underlying the phytoplankton patchiness formation[4], since the latter has profound effects on the ecology of the oceans [5]. These patterns play a fundamental role in microorganisms populations composition, modulating cells activities like the encounter rate, the predation and the reproduction [3]. Here we focus on mechanical stresses, in particular on how cells behave within a turbulent flow. It is known that motile microorganisms can produce patchiness in presence of a turbulent flow [2] but a largely unanswered question concerns the responses of non-motile cells to environmental signals. Motivated by the physiological regulation of buoyancy prevalent in non-motile phytoplankton species [1], we investigate, by means of direct numerical simulations (DNS), the dynamics of cells whose buoyancy depends on local strain rate within a three- dimensional turbulent flow. Particles dynamics is parametrized by the sedimentation number Π = vmax/uη and the strain rate number Σ = SH τη (where vmax is the maximum sedimentation velocity, SH is the strain rate constant and uη , τη are the Kolmogorov velocity and time scale). In particular we consider two possible kinds of particles, differing in the sign of the response: we refer to cells whose density decreases (increases) with the mechanical stresses as shear-thinning (shear-thickening) respectively. With our approach we show that, in contrast to passive tracers, the buoyancy regulation strategy leads to clustering formation, we demonstrate how its intensity depends on settling speed (fig.1), how different cells sample different flow regions and finally we discuss how sedimentation time of active cells differs form which of passive ones. 16:30 15 mins #422 SETTLING DYNAMICS OF INERTIAL PARTICLE David De Souza, Romain Monchaux, Anne Dejoan Abstract: Dispersed two phase flows are relevant to many fields, as they occur in both natural phenomena (rain formation, marinesnow) and technological applications (combustion chambers, chemical reactors). Such flows are generally described using the following parameters : the Reynolds number (turbulence intensity), the Stokes number (particle inertia), the Rouse number and the Froude number (effects of gravity). Given the complexity of the equation of a particle moving in a turbulent flow [2, 4], most models rely on heavy simplifications, which is why further experiments are still needed. Among the reported behaviours of these flows, preferential concentration and settling speed alteration of inertial particles when the carrying phase is in a turbulent state have largely been documented. Moreover, the tendency of the dispersed phase to preferentially accumulate in certain regions of the flow while leaving others completely void seems to be connected to the modification of its settling speed, as observed in the experiments of Aliseda [1]. Aliseda suggested that as the particles coalesce into clusters they would form bigger meta-particles that would be responsible for the alteration of the settling velocity. This claim is supported by recent experiments [3] and numerical works involving a back reaction of the fluid phase on the particles [5]. In these two-way simulations, particles and fluid have been observed to fall together, which is consistent with the meta particles hypothesis. To study these phenomena, we devised an experiment in which heavy solid particles fall in a turbulent flow generated by oscillating grids. The goals of our study are the following (i) to disentangle the effects of the different parameters and (ii) to further probe the relation between local particle/fluid slip velocity and settling speed modification. By tuning our particle populations in both density (using glass, ceramic, steel and tungsten carbide particles giving us the following particle/water density ratios 2.5, 4, 7 and 14) and size (10 to 200 μm refined using a set of sieves) we can map our parameter space to assess the influence of each one relative to the others. Moreover, the use of a double measurement setup with particle image velocimetry (PIV) and particle tracking (PTV) gives us access to both fluid and particle velocities, enabling us to test the meta-particle hypothesis. To complement this experimental study, further two-way numerical simulations are also being conducted. As a first step, we studied the settling of particles in a quiescent fluid, with particle volume fractions high enough for collective effect to be expected (typically 10−5 to 10−4). We compared these results to settling speeds expected using different drag models (Stokes, Schiller-Naumann, Newtonian...). In the litterature, when discussing the alteration of the settling velocity, the reference velocity used by most authors is that of a particle subject to linear Stokes drag which assumes a still fluid phase. As collective effects occur even without turbulence, using this settling model as a reference can be detrimental when trying to specifically assess the effect of upstream turbulence on the settling velocity, especially when studying potential hindering effects. For further studies, we advise the use of reference measurements, like those presented here. Our next step is to compare our previous measurements with experimental and numerical results obtained when turbulence is added. References [1] A. Aliseda, A. Cartellier, F. Hainaux, and J. C. Lasheras. Effect of preferential concentration on the settling velocity of heavy particles in homogeneous isotropic turbulence.J. Fluid Mech.,468:77–105, 2002. [2] Renée Gatignol. The Faxén formulae for a rigid particle in an unsteady non uniform Stokes flow. J. Mécanique Théorique Appliquée,1(2):143–150, 1983. [3] P. D. Huck, C. Bateson, R. Volk, A. Cartellier, M. Bourgoin, and A. Aliseda. The role of collective effects on settling velocity enhancement for inertial particles in turbulence.J. Fluid Mech.,846:1059–1075, 2018. [4] Martin R. Maxey and James J. Riley. Equation of motion for a small rigid sphere in a non uniform flow.Phys. Fluids, 26(4):883–889, 1983. [5] R. Monchaux and A. Dejoan. Settling velocity and preferential concentration of heavy particles under two-way coupling effects in homogeneous turbulence. Phys. Rev. Fluids,2(10), 2017. 16:45 15 mins #552 Turbophoresis of small heavy particles in homogeneous turbulence Robin Vallee, Jérémie Bec Abstract: Small heavy particles transported by a turbulent flow detach from the flow and form uneven distribution leading to clustering or preferential concentration. That fundamental phenomenon, called turbophoresis, is mainly observed in inhomogeneous turbulent flow where particles concentrate in low energy regions. We show here that this phenomenon can also be observed in homogeneous turbulent flows and we illustrate and quantify the importance of instantaneous inhomogeneities by performing direct numerical simulations. 17:00 15 mins #156 Broadening of cloud droplet size distribution and liquid water content spectrum in turbulence Izumi Saito, Tatsuya Yasuda, Toshiyuki Gotoh, Takeshi Watanabe Abstract: We conducted large-scale direct numerical simulations (DNSs) of cloud droplets in turbulence using “cloud microphysics simulator", a DNS model for cloud turbulence which uses the Lagrangian dynamics for the droplets and the Eulerian dynamics for the turbulent flow fields [2]. Motivated by the recent studies by the “Π-chamber" [1], the laboratory cloud chamber experiments, we performed the numerical experiments in the following setups. Initially unactivated cloud droplets are injected into the air turbulence in a cubic periodic box with a side L box = 1.024m at a constant rate, removed from the box with a residence time τ res = 580s (according to the Poisson process), so that the steady state of the particle number density is achieved through the balance between the source and sink. Cloud droplets are advected in the box and change their sizes by condensation/evaporation, where the fluctuations in turbulent air velocity and supersaturation fields are excited by external random force and source. Cloud droplets are assumed to be non-inertial, and gravity and collision-coalescence processes are not included. The total number of grid points for fluid is N = 512^3 , and the Taylor microscale Reynolds number is R_λ = 207. Fig. 1a shows the size distributions of cloud droplets at statistically steady states, where the peak shifts toward smaller sizes for runs with increasing droplet number densities (n d = 20, 80, 200, 500 [cm^{-3} ] for Runs 1–4, respectively). This tendency is associated with the dispersion aerosol indirect effect. The results for Run 1–3 are quantitatively consistent with those for the Π-chamber. We also investigated the spectra of the cloud water mixing ratio (liquid water content, LWC) and the effects of gravity and particle inertia on those spectra. Fig. 1b shows the LWC spectra for Run 3 with (black curve) and without (green curve) the effects of gravity and particle inertia. As shown in the previous study [3], the LWC spectrum consists of two parts, the one from the uniform random distribution of particles (uncorrelated part, proportional to k 2 ) and the deviation from the uncorrelated part (correlated part). Fig. 1b shows the correlated part of LWC spectra obtained by subtracting the uncorrelated part from the total LWC spectrum as in [3]. Without the effects of gravity and particle inertia (green curve), the LWC spectrum has the slope close to -1 for higher wavenumbers (k ≥ 10), and the slightly steeper slope for lower wavenumbers (k < 10). On the other hand, with the effects of gravity and particle inertia (black curve), the LWC spectrum shows a bump for higher wavenumbers (k ≥ 10). Remarkably, such a LWC spectrum with a bump resembles the LWC power spectrum observed in the mountaintop clouds [4]. We will report on the formation mechanism of these LWC spectra in our DNS and also provide results for DNSs with even higher Reynolds numbers. References [1] K. K. Chandrakar, W. Cantrell, K. Chang, D. Ciochetto, D. Niedermeier, M. Ovchinnikov, R. A. Shaw, and F. Yang. Aerosol indirect effect from turbulence-induced broadening of cloud-droplet size distributions. Proc. Nat. Acad. Sci., 113:14243–14248, 2016. [2] T. Gotoh, T. Suehiro, and I. Saito. Continuous growth of cloud droplets in cumulus cloud. New J. Phys., 18:043042, 2016. [3] I. Saito and T. Gotoh. Turbulence and cloud droplets in cumulus clouds. New J. Phys., 20:023001, 2018. [4] H. Siebert, R. A. Shaw, J. Ditas, T. Schmeissner, S. P. Malinowski, E. Bodenschatz, and H. Xu. High-resolution measurement of cloud microphysics and turbulence at a mountaintop station. Atmos. Meas. Tech., 8:3219–3228, 2015. 17:15 15 mins #331 Transport properties of quasi-neutrally-buoyant inertial particles Marco Martins Afonso, Sílvio M.A. Gama, Andrea Mazzino, Paolo Muratore-Ginanneschi Abstract: We investigate the sedimentation and diffusion of quasi-neutrally buoyant inertial particles carried by incompressible zero-mean fluid flows. We obtain generic formulae for the terminal velocity and the effective diffusivity in generic space-and-time periodic (or steady) flows, along with further information for flows endowed with some degree of spatial symmetry such as odd parity in the vertical direction, and for parallel flows as well. These expressions consist in space-time integrals of auxiliary quantities which satisfy partial differential equations of the advection-diffusion-reaction type, that can be solved at least numerically since our scheme implies a huge reduction of the problem dimensionality from the full phase space to the classical physical space. The discrepancy in mass density between fluid and particles is assumed to be small, and represents the basic parameter for a regular perturbative expansion. By means of analytical techniques such as Hermitianization, our procedure provides results also for finite particle inertia, away from the over-damped limit of quasi-tracer dynamics. In the same framework, we also analyze the evolution of the distribution of inertial particles released by a spatially-localized (punctual) source, such as a chimney emitting some pollutant or a syringe injecting into a microchannel, at a constant rate. We will show results for this quantity and for its physical-space counterpart as functions of the emission distribution in the covelocity space. Namely, the parity and the overall integral of the latter quantity are key factors in determining the order in our expansion at which the leading correction takes place. 17:30 15 mins #514 Design, construction and characterization of instrumented particles for the Lagrangian characterization of turbulent flows. Facundo Cabrera, Pablo Cobelli Abstract: I will present the design, construction and characterization of a spherical particle (diameter 36 mm) for the Lagrangian characterization of turbulent flows. This is an autonomous local measurement equipment, capable of registering its own translational acceleration and angular velocity components in three dimensions. Instrumented particle and PTV were used to characterise the Lagrangian dynamics of this instrumented particles in the gravity wave turbulence regime, focusing both on translational and rotational dynamics. 17:45 15 mins #339 SPHEROIDS IN DECAYING TURBULENCE FROM TAYLOR-GREEN VORTEX FLOW Rohith Jayaram, Jurriaan Gillissen, Lihao Zhao, Helge I Andersson Abstract: In our work, decaying turbulence is studied in the context of Taylor-Green vortex (TGV) flow. As time progresses, coherent structures will break down leading to 3D turbulence. Adding particles to this flow will be very interesting to explore the particle dynamics in relation to the flow. Generally, spherical (isotropic) particles are studied in the flow owing to its ease in terms of numerics and mathematics. Here, we are also considering non-spherical particles and therefore an intimate coupling between the translational and rotational motion is investigated. Non-spherical particles are approximated as axisymmetric ellipsoidal particles, i.e prolate (rod-like) and oblate (disk-like) spheroids. Direct numerical simulations of the flow field coupled with a Lagrangian point-particle tracking methodology is employed. Our study extends to explore the effects of particle inertia, shape and fluid shear on orientational and rotational dynamics of spherical and spheroidal particles. 18:00 15 mins #290 PATH-PLANNING SMART SWIMMERS IN TURBULENT FLOWS Rahul Pandit, A. Jaya Kumar, Akhilesh Kumar Verma, Jeremie Bec Abstract: Path planning in a dynamic environment is a hard problem. We employ adversarial-reinforcement learning to find nontrivial paths and policies, to outperform naïve strategies for passive-smart swimmers in two-dimensional (2D) and threedimensional (3D) statistically isotropic and homogeneous turbulent flows, which we study via pseudospectral direct numerical simulations (DNSs) with periodic boundary conditions. We introduce passive smart swimmers into the flow at random positions and with random orientations. Over a period of time, these swimmers learn how to control their directions of motion, and optimise their paths (in terms of the time of arrival). We perform numerical experiments to demonstrate such improvement in performance; we also analyze the dependence of this improvement on smart-swimmer control parameters. Note that it is not obvious, a priori, if there exists a stable, non-trivial, optimal strategy, for such swimmers in a turbulent flow, that outperforms the naïve strategy in which the swimmer points along the displacement vector from its current position to its target point.