Settling of inertial particles in turbulent Rayleigh-Benard convection

Vojtech Patocka; Enrico Calzavarini; Nicola Tosi

doi:10.1103/PhysRevFluids.5.114304

Patocka, Vojtech;Calzavarini, Enrico;Tosi, Nicola

2020-05-11 00:00:00

1, 2 1 Vojt ech Pato cka, Enrico Calzavarini, and Nicola Tosi Institute for Planetary Research, German Aerospace Center (DLR), Berlin, Germany Univ. Lille, ULR 7512 - Unit e de M ecanique de Lille - Joseph Boussinesq (UML), F-59000 Lille, France The settling behaviour of small inertial particles in turbulent convection is a fundamental problem across several disciplines, from geophysics to metallurgy. In a geophysical context, the settling of dense crystals controls the mode of solidi cation of magma chambers and planetary-scale magma oceans, while rising of light bubbles of volatiles drives volcanic outgassing and the formation of primordial atmospheres. Motivated by these geophysical systems, we perform a systematic numerical study on the settling rate of particles in a rectangular two-dimensional Rayleigh-B enard system with Rayleigh number up to 10 and Prandtl number from 10 to 50. Under the idealized condition of spherically-shaped particles with small Reynolds number, two limiting behaviours exist for the settling velocity. On the one hand, Stokes' law applies to particles with small but nite response time, leading to a constant settling rate. On the other hand, particles with a vanishing response time are expected to settle at an exponential rate. Based on our simulations, we present a new physical model that bridges the gap between the above limiting behaviours by describing the sedimentation of inertial particles as a random process with two key components: i) the transport of particles from vigorously convecting regions into sluggish, low-velocity \piles" that naturally develop at the horizontal boundaries of the system, and ii) the probability that particles escape such low-velocity regions without settling at their base. In addition, we identify four distinct settling regimes and analyze the horizontal distribution of sedimented particles. For two of these regimes settling is particularly slow and the distribution is strongly non-uniform, with dense particles being deposited preferentially below major clusters of upwellings. Finally, we apply our results to the crystallization of a magma ocean. Our prediction of the characteristic settling times is consistent with fractional crystallization, i.e. with the ecient separation of dense crystals from the residual lighter uid. In absence of an ecient mechanism to re-entrain settled particles, equilibrium crystallization appears possible only for particles with extremely small density contrasts. Keywords: Particle settling; Rayleigh-B enard convection; Numerical modelling; Magma ocean; I. INTRODUCTION The motion and sedimentation of small particles in a convecting uid is of great interest to uid dynami- cists, geophysicists, as well as to metallurgists. Dispersion of pollutants, dust [1], and organic material such as pollen in the atmosphere [2], phytoplankton population dynamics in oceans and lakes [3, 4], crystallization of magma chambers [5, 6], of primordial magma oceans in rocky planets [e.g. 7, 8], and of planetary metallic cores [e.g. 9, 10] are just a few examples of geophysical processes that to some extent can be described through the settling of inertial particles in a uid undergoing highly vigorous convection. In industrial settings, the puri cation of alloys [11] and micro uidic heat transfer technologies [e.g. 12] count into this group. From a general point of view, the phenomenology of particle-laden turbulent ows has been the subject of extensive studies over the last decades. Researches have focused on the statistical characterisation of particles' dispersion and accumulation upon varying the ow turbulence intensity, the particles' inertia (which is linked to their size and mass density), and their shape [13, 14]. For what concerns particle turbulent settling, there are nowadays solid numerical and experimental evidences that particle sinking is enhanced by the eect of turbulence, while the opposite happens for particle rising [for a review, see 15]. However, these results have been obtained in the context of idealized ows: kinematic ows [16, 17], unbounded turbulence [18], or channel and pipe ow geometries. Much less explored is the context of thermally-driven ows [for experimental studies, see 5, 19]. Despite the relevance of the problem of particles settling in turbulent ows for a variety of natural and industrial systems, this work is largely motivated by the study of the crystallization of primordial magma oceans, with our choice of the parameter space being inspired by this system (Section II A). Planetary-scale patocka.vojtech@gmail.com arXiv:2005.05448v1 [physics.flu-dyn] 11 May 2020 Settling of particles in thermal convection Submitted to Phys. Rev. Fluids volumes of liquid silicates are thought to form during the accretion and dierentiation of terrestrial bodies like the Earth, Mercury, Venus, Mars and the Moon [e.g., 20{22]. The way magma oceans solidify is of primary importance for the long-term thermo-chemical evolution of the interior of planets [e.g., 23]. Whether or not newly formed crystals settle or remain suspended by turbulent ow determines the initial distribution of the composition of the silicate mantle [e.g., 7]. This is a dicult problem that depends on the density contrast between crystals and melt, the melt viscosity, the size of the crystals and the convective dynamics of the system. If dense crystals are eciently maintained in suspension, a magma ocean undergoes equilibrium (or batch) crystallization, which leads to a largely homogeneous composition of the rocky mantle. By contrast, if crystals tend to settle and crystal-melt separation is ecient, fractional crystallization takes place. Residual melts are progressively more and more enriched in so-called incompatible elements such as iron-oxides and heat-producing elements. This process ultimately leads to a compositionally-strati ed mantle whose long- term evolution can be dramatically dierent from that of a homogeneous one [e.g., 24{27]. Magma chambers are small-scale analogs of magma oceans. Upon cooling and solidi cation these typically undergo strong fractionation, which is often attributed to crystal settling [e.g., 5, 6, 28, 29]. Similar to the settling of negatively buoyant crystals, oating of light particles is also a fundamental process in the context of the crystallization of magma oceans and magma chambers. In fact, it is the basic mechanism underlying magma degassing, where gas bubbles are released from volatile-saturated magma [e.g., 30]. Greenhouse volatiles such as H O and CO also behave as incompatible species and tend to 2 2 be strongly enriched in the liquid phase upon magma crystallization. The eciency with which these are released from a magma ocean controls the formation of primordial atmospheres and the timescale of magma ocean solidi cation [e.g. 31{33]. In the context of a crystallizing magma, Marsh and Maxey [28] modelled the transport of particles by convective motions as a turbulent diusion process, which was a common approach in studies of mixing in turbulent ows [e.g. 34, 35]. Fundamental laboratory experiments aimed at assessing settling rates in a cooling magma were later performed by Martin and Nokes [5], who employed the turbulent diusion theory to explain their measurements. Assuming the concentration of particles to be spatially uniform, they derived a simple model for particles with a vanishing Stokes velocity according to which the number of suspended particles decays exponentially with time. Although Martin and Nokes [5] anticipated that for particles with a larger Stokes velocity their \diusion model of turbulent transport will begin to break down and other assumptions will no longer be valid, in particular the assumption of one-dimensionality", surprisingly little eort has been devoted to extend their work. To our knowledge, no experimental or numerical study has been performed that systematically explores the settling mechanism of particles with a non-vanishing Stokes velocity in turbulent, thermally-driven convection. Dierentiation of a cooling magma is a competitive process between generation, sedimentation and re- entrainment of crystals. The problem of re-entrainment, in particular, has been addressed by various authors both theoretically [e.g., 36] and experimentally [19, 36]. Although the lifting of negatively buoyant particles from the crests of dunes has been recognized as one of the main mechanisms to keep these in suspension [36], in this study we focus entirely on the settling process and completely neglect re-entrainment, which we plan to address in future work. As soon as our particles reach the bottom boundary of the domain (resp. the top boundary for light particles), we eliminate them from the ow, not allowing any accumulation or subsequent lifting of the sedimented material. We use a modeling approach based on a Eulerian-Lagrangian description of the uid ow and the particu- late phase respectively, and track individual trajectory of each particle. Our approach thus captures how the exact ow structure aects the particle motion. This level of detail is still challenging from the experimental point of view and it brings new results when compared to the one-dimensional turbulent diusion theory. For example, horizontal variations in the distribution of sedimented particles can be evaluated and linked to the large scale circulation of the uid (also called the \wind of turbulence", see e.g. Ahlers et al. [37] for a discussion). Based on the experimental work of Koyaguchi et al. [6], Sparks et al. [38] argued for cyclic sedimentation of crystals in magma chambers caused by the cessation of convection due to the particle concentration exceeding a certain critical value. Similar behaviour was observed by H oink et al. [39] in the context of numerical simulations of metal-silicate separation, and by Verhoeven and Schmalzl [40] in a numerical study that combines a nite volume convection code with a discrete element method [41]. Here we neglect the in uence of particles on the convective ow [e.g. 42], i.e. we assume only small particle concentrations. We believe that the dynamics of dilute suspensions is suciently rich to warrant a study entirely dedicated to particle settling before considering additional complexities arising from larger solid fractions. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids For particles with small radii Verhoeven and Schmalzl [40] obtain statistically stationary suspension in which convective motions keep particles inde nitely entrained. Similar results were obtained in the non- rotating cases of Maas and Hansen [43, 44] whose model builds on the one by Verhoeven and Schmalzl [40]. Although it is at odds with the early results of Martin and Nokes [5], Maas and Hansen [44] conclude that \it is generally assumed that vigorous convection would prevent major gravitational segregation in a magma ocean at all latitudes (Andrault et al., 2017)". Here we employ a more elaborate particle model and refute such statement, con rming the experimental results of Martin and Nokes [5] in which particles with a vanishing response time settle at an exponential rate. The rest of the paper is organized as follows. In Section II we introduce our numerical model and discuss the choice of model parameters. In Section III we present the settling curves of a reference simulation and classify them according to four distinct regimes. In Section III D we then introduce a general model that describes particle settling as a random process. In Section IV A we discuss the horizontal distribution of sedimented particles, showing how it can be strongly non-uniform in some regimes. In Section IV B the focus is on particles lighter than the uid, including bubbles. These become concentrated in ow vortices, which signi cantly delays their rising. In Section IV C we analyze how our results depend on the strength of convective vigor and uid inertia. Finally, in Section V the results are extrapolated to the environment of an extremely vigorous, global magma ocean and that of a large magma chamber. II. GOVERNING EQUATIONS Rayleigh-B enard convection of an incompressible isoviscous uid is governed by the Boussinesq equations: @ U + (Ur)U = rP= + r U (T T )g; (1) 0 0 r U = 0; (2) @ T + (Ur)T = r T; (3) where U(X; ) and T (X; ) are respectively the velocity and temperature elds, the kinematic viscosity of the uid, the mean mass density at the reference temperature T , the volumetric thermal expansion 0 0 coecient with respect to the reference temperature, g the gravitational acceleration, the thermal diu- sivity. The hydrostatic stress, rP = g, is already subtracted from Eq. (1), leaving only the dynamic 0 0 pressure P on its right-hand side (RHS). As shown by the last term in Eq. (1), only temperature-induced variations of density are considered to drive the ow. Equations (1) { (3) are solved in a 2D box with periodic side walls and aspect ratio 2. No-slip conditions are assumed on the top and bottom boundaries, which are are isothermal, with a constant temperature dierence T driving thermal convection. We non-dimensionalize the governing equations by scaling the length with the height of the box H , x := X=H , the velocity with the characteristic velocity u := gTH , u := U=u , the density with the reference density , and we introduce non-dimensional temperature := (T T )=T . For the time and 0 0 pressure it then follows: t := u =H , and p := P=( u ). In terms of non-dimensional quantities, the governing equations read: Pr @ u + (ur)u = rp + r u + ^z (4) Ra r u = 0 (5) @ + (ur) = p r ; (6) PrRa where Ra and Pr are the Rayleigh and Prandtl number that control the ow characteristics: gTH Ra := ; Pr := : (7) The uid carries inertial particles, whose trajectory is governed by friction from the surrounding uid in combination with particle buoyancy. Under idealized conditions of spherically-shaped particles with small Reynolds number, the Lagrangian equation of motion for a massive particle reads [e.g. 45]: dV DU 1 = + (U V) + (1 )g; (8) d Dt D Settling of particles in thermal convection Submitted to Phys. Rev. Fluids where V is the particle velocity and the rst term on the RHS denotes the material derivative of the uid velocity. The modi ed density ratio = 3 =( + 2 ) relates the density of the uid with the particle f f p f density . Eq. (8) is a truncated version of the original equation derived independently by Maxey and Riley [46] and Gatignol [47]. Due to the small Reynolds number and size of the particles we neglect here both the unsteady drag term, known as history term, and the Fax en corrections. We systematically vary the ratio = and assume that it is constant for each particle throughout the f p simulation. This assumption does not necessarily neglect the density variations of material: in view of the Boussinesq approximation = [1 (T T )], employed in Eq. (1), the assumption simply means f 0 0 that each particle is always at the same temperature as the surrounding uid and has the same thermal expansivity [for eects resulting from keeping the particles at a dierent temperature than the uid, see 48]. The particle response time = r =(3 ) depends quadratically on the particle radius r , which we also D c c vary systematically. After non-dimensionalizing it with the same scales introduced above, Eq. (8) takes the form: dv Du 1 = + (u v) + ^z; (9) dt Dt St leaving three non-dimensional parameters to control the particle dynamics, namely: 3 r gT 1 = ; St = ; = : (10) + 2 T f p 3 H The Stokes number St is the particle response time divided by the characteristic time H=u . It describes the viscous friction acting on each particle due to the dierence between particle and uid velocity. The parameter (hereafter buoyancy ratio) expresses the relative importance of particle buoyancy with respect to the thermally-induced buoyancy of the uid (the unit vector z ^ points vertically upward). The rst term on the RHS of Eq. (9) is the so called added-mass as estimated by [49]. We do not consider any feedback mechanism with respect to the ow: the uid velocity u is obtained from Eqs. (4){(6) and does not depend on the particle velocity v, i.e. we adopt a one-way coupling [for the distinction between one-way and two-way coupling, see the review of 50]. In a turbulent ow, the adopted particle model can be considered appropriate as long as the particle size, r (resp. r =H in dimensionless units) is up to the same order of magnitude as the spatial dissipative scale c c of turbulence, . In Rayleigh-B enard ow the global value of such scale, in the current dimensionless units, 1=2 1=4 goes as = Pr (Ra(Nu 1)) [e.g. 51], meaning that it decreases at increasing the thermal forcing Ra (and so the Nusselt number Nu) but increases at increasing the Prandtl number Pr (see also Discussion). For particles suspended in a uid at rest, i.e. with u 0, Eq. (9) can be solved analytically, yielding: v ~(x; t) = v (x) exp + St z ^: (11) St In the limit t!1, the so-called terminal or Stokes' velocity v is reached: 2 r g f p v ~(t!1) = St z ^ = z ^ =: v z ^; (12) 9 u where the terminal velocity is de ned positive for sinking particles and negative for rising particles. We inject 10 particles of 301 dierent types into a fully developed, two-dimensional, statistically-steady thermal convection, with each particle type represented by three values: St, , and . Since we are primarily interested in the dynamics of the particles, we refer to the thermal ow of the carrier as the \background" ow. For particles with v > 0 (, < 0 , < 1), i.e. those denser than the uid (labeled as heavy), we measure the time it takes until they reach the bottom boundary. For particles with v < 0 (, > 0 , > 1, labeled as light) we do the same with respect to the top boundary. For brevity, both these cases are referred to as \settling". Initially, all particles are distributed uniformly across the domain and their velocity is set equal to the local velocity of the uid. 300 dierent types of particles are obtained by evenly sampling = f p and r ; one particle type is reserved for uid tracers. The above described model system is numerically simulated by means of the Eulerian-Lagrangian code ch4-project [52]. The code adopts a lattice Boltzmann (LB) algorithm for the computation of the uid and temperature dynamics, while it uses a second order time-stepping and grid-to-particle bi-linear interpolation for the computation of particles' trajectories. This code has been already extensively employed in studies involving turbulent thermal convection and inertial particle dynamics [53]. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids A. Model parameters We aim to map the settling behaviour of particles over the entire St; , and space, while focusing on 8 10 12 highly vigorous convection (Ra = [10 ; 10 ; 10 ]), with moderate to small importance of inertia (Pr = [10; 50]). As such, our results are applicable to a range of natural systems (see Section I). Throughout the paper we strictly use non-dimensional control parameters, but it is instructive to demonstrate how these are linked to physical parameters of a particular system, namely the thermal convection of a large reservoir of crystallizing magma. In this section we inspect how the parameter space is mapped and discuss intrinsic limitations of our numerical approach. In Table I we list the physical parameters that roughly describe the thermal convection of a primordial, mantle-deep magma ocean for the Earth. A relativley large uncertainty is in the value of the viscosity of high-pressure and -temperature magma. First-principles simulations suggest that the kinematic viscosity of MgSiO , one of the major mantle silicates, over the temperature and pressure range relevant for a global 5 6 2 magma ocean ( 2000 4000 K and 0 130 GPa) is on average of the order of 10 10 m /s [54]. Since the Prandtl number is de ned as =, the lower and upper bounds of de ne the range of interest of Pr and we indicate directly as Pr in Table I. The temperature contrast T driving convection is also dicult to determine precisely. The reference value of only 1 K reported in the table roughly corresponds to the contrast predicted by parameterized models of the thermal evolution of the Earth's magma ocean in the presence of an atmosphere [e.g., 32, 33]. Such a low value is also representative for planetary cores, where a large volume of low-viscosity metallic liquid undergoes thermal convection [e.g., 55]. TABLE I. Parameters of a global, mantle-deep magma ocean Parameter Symbol Value Units Mantle depth H 2890 km Reference grav. acceleration g 9:8 m/s ref a 5 1 Thermal expansivity 510 K b 7 2 Thermal diusivity 510 m /s c 2 Kinematic viscosity [10; 50] m /s Temperature contrast T 1 K a ref Crystal size r [0:5; 10] mm Density ratio = [0; 2] { p f a b c from Solomatov [8]; from Ni et al. [56]; see Karki and Stixrude [54] for typical viscosities of silicate liquids at high pressure and temperature; see e.g. Lebrun et al. [32] and Nikolaou et al. [33] for typical temperature contrasts during the evolution of magma oceans. ref Sampling the ranges of r and = from Table I results in a sampling of the non-dimensional parameter p f space ; St; (set A in Fig. 1). The y-axis in Fig. 1 represents the absolute value jj rather than in order to t both light ( > 0) and heavy ( < 0) particles into a compact plot. The modi ed density ratio is marked by color only. Later we will show that, apart from the eect described in Section IV B, the rst term on the RHS of Eq. (9) has secondary importance on the settling behaviour, which sidelines the relevance of The terminal velocity v and the response time St can be used to a priori estimate the number of time steps that are required to evaluate the settling time of a given particle type. In the time 1=jv j, each particle would cross the model domain vertically if sinking (or rising) at the speed v , making 1=jv j a proxy for t t the minimum required duration of a simulation (and thus CPU time). Due to constraints arising from the numerical integration of Eq. (9), which we explain below, we use max(10t=St; 1)=jv j to estimate the minimum required CPU time (green line in Fig. 1). Here, t denotes the maximum time step allowed by the Courant-Friedrichs-Lewy condition, i.e. by the advective and diusive time scales of the background ow. For illustration purposes, let us assume u 0 and discretize Eq. (9) via explicit, rst-order Euler scheme: t t v ~ = v ~ 1 + v : (13) n+1 n t St St It follows that t=St must be smaller than 2 in order to avoid numerically unstable solutions. Demanding numerical accuracy limits the admissible values of t=St even further { only a small fraction of v must be t Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 1. Model parameter space describing particle dynamics. Red dashed and solid lines show three isolines of the terminal velocity v . Colored symbols show three dierent simulation sets: A, B, and C (triangles, circles, and squares). The green line marks a constant value of required CPU time. We performed numerical simulations corresponding only to the sets B and C (see text for details). Pr is equal to 50. added at each time step to ensure that convergence to the stationary solution v is smooth. In a turbulent ow (u 6= 0), Eq. (9) yields accurate trajectories only when t=St < 0:1 (the exact value depends on the employed numerical scheme, with 0.1 resulting from our experience with the second-order Adams-Bashforth formula that we use to advect the particles). This constraint increases the CPU time of each simulation (resp. decreases the allowed time step) by an additional factor, 10t=St, where St is the smallest Stokes number in the respective set of particles. Red lines in Fig. 1 mark isolines of v . In Section III, we show that to rst-order the settling behaviour of particles can be described by their terminal velocity only: for a given background ow, particles with the same v settle in a similar manner. Since the green and red lines in Fig. 1 have dierent slopes, it is convenient to modify the parameters from Table I to move along the isolines of v in the direction of smaller CPU time (i.e. to the right of the St;jj space). Sets B and C in Fig. 1 are two such modi cations of the ref ref original set A, obtained by setting: B) r = h2; 20i cm, = = h0:99; 1:01i, and C) r = h1; 10i cm, p f c c = = h0; 2i, = 2 10 , T = 1000 K. While set A is computationally dicult to reach and would p f require close to a year on several hundreds of CPU cores, sets B and C can be completed within a month on a 32-core machine. For the purpose of this study, particle sets B and C can be simply understood as the selected coverage of model parameter space (later we argue that based on these sets we map the entire St; ; space reasonably well). We note, however, that both sets also have a certain geophysical interpretation. When compared Settling of particles in thermal convection Submitted to Phys. Rev. Fluids to the original set A, set B has enlarged r and a narrowed = range. As such, its parameters roughly c p f correspond to large clusters of crystals with density similar to that of the surrounding magma. Set C has a reduced thermal expansivity and a larger temperature contrast that lies in the range of temperature contrasts characteristic of magma oceans that cool in absence of an atmosphere [32, 33]. Based on the parameter values listed in Table I, the Rayleigh number Ra of a mantle-deep magma ocean ref would be of the order of 10 . This is far from being reachable with any numerical method because the 1=3 thickness of the thermal boundary layer in a convecting system scales approximately as Ra , demanding higher resolution for higher Ra. Here we model a series of Rayleigh numbers up to Ra = 10 , using up to 40962048 grid points. The possibility of extrapolating our results to higher Rayleigh numbers is analyzed in Section V. We note that our code is based on a dimensional formulation. Therefore, in order to reduce the Rayleigh number we need to modify some of the parameters in Table I. Our aim is to modify the parameters such as to change Ra and leave the remaining control parameters Pr; St; , and untouched, regardless of the choice of Ra. This can be achieved by replacing g with a reduced gravitational acceleration, g := g Ra=Ra = ref ref ref 3 ref ref 1=4 Ra=(TH ), and by replacing r with an in ated crystal size, r := r (g =g) . In this way, the c ref c c coverage of the St; ; space remains identical for all tested values of Ra (i.e. Fig. 1 remains the same regardless of the value of Ra). For each simulation set, we rst wait for thermal convection to develop into a statistically steady state and then we inject all the particles at once, distributing them uniformly in space and assigning them the velocity of the carrier uid, i.e. v(t=0; x; z) = u(t=0; x; z). In Fig. 2a, we show the average root mean square velocity of the background ow for all the tested values of Ra and Pr. Fig. 2b shows the corresponding Reynolds 8 10 number, Re := U H=. For Pr = 10 we run simulations with Ra equal to 10 and 10 , while for Pr = 50 rms 8 10 12 we test three values, Ra = 10 , 10 , and 10 (lowering Prandtl number increases the resolution demands { see the Reynolds number for two simulations with the same Ra but dierent Pr). Our simulation sets are labeled as B or C, depending on the range of particle parameters (see Fig. 1), and by upper and lower indices 10 10 we label the exponent of Ra and the value of Pr (e.g., C stands for simulation set C with Ra = 10 and Pr = 50). 8 10 12 8 10 FIG. 2. (a) Volumetric averages of the root mean square velocities in simulation sets C , C , C , C , and C . 50 50 50 10 10 The number in parenthesis indicates the time-averaged value u that is used later for computing the v =u ratio. rms t rms (b) The corresponding Reynolds number, U H=. rms Settling of particles in thermal convection Submitted to Phys. Rev. Fluids III. RESULTS: SETTLING CURVES In Section II A we anticipated that the terminal velocity v is capable of sorting the settling behaviour of particles in a ow of given Ra and Pr. In order to compare the settling behaviour also across ows with a dierent convective vigor, one more parameter is needed. Similarly to the experimental study of Martin and Nokes [5], who divide v by the average vertical velocity of the ow and use the resulting ratio to organize their results, we use u to account for the properties of the background ow. In this section, we show rms that there are four distinct regimes describing the sinking or rising of particles, and that the ratio jv j=u t rms determines to which regime a given particle type belongs. In Fig. 3 we plot the temporal evolution of the settling process for the simulation set C , which is taken as a reference case. When heavy particles (v >0) reach the bottom, resp. light particles (v <0) reach the t t top, we eliminate them from the ow and mark them as settled. Each line in Fig. 3 represents a dierent particle type, although we do not show the respective values of St; ; and . Instead, we mark each line by the value of jv j=u { this single parameter uniquely orders the obtained settling curves. We identify four t rms distinct groups: i) \stone-like" regime, jv j=u > 2:0 (Fig. 3a); ii) bi-linear regime, 0:3 < jv j=u < 2:0 t rms t rms (Fig. 3b); iii) transitional regime, 0:02 < jv j=u < 0:3 (Fig. 3c); and iv) \dust-like" regime jv j=u < 0:02 t rms t rms (Fig. 3d). The time on the x-axis of Fig. 3 is multiplied by v for each settling curve individually. The x-axis thus represents distance rather than time. The \terminal distance", t v , corresponds to the distance a particle with a given v would travel in a uid at rest by the time t (i.e. t v = 1 represents sinking with the Stokes t t velocity through the entire container). This means that, even though the particles in Fig. 3d seemingly take only approximately 5 times longer than those in Fig. 3a to completely settle, the actual time diers by more than two orders of magnitude because the corresponding value of v diers by more than a factor hundred in both subplots. The same applies for dierent settling curves within each subplot: two settling curves that overlap but have dierent colors correspond to dierent settling rates with respect to time t. FIG. 3. Settling curves obtained from the simulation set C . All particle types are separated into four groups based on the jv j=u ratio, which is shown in color for each subplot. The x-axis represents the terminal distance t v . t rms t Black and red solid lines represent the analytic solutions (14) and (15) respectively. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids The black line in Fig. 3 is the theoretical prediction N t = jv~j dt = jv jtjv jSt 1 exp (14) t t N St 0 0 where N and N are the number of settled particles and the initial number of particles respectively. The s 0 velocity v~ is given by Eq. (11), and zero initial conditions are considered, v~(x; t=0) 0. Eq. (14) thus expresses the percentage of particles that would settle at the time t if initially they were distributed uniformly in a still uid. The shape of the black curve is nearly identical to simply min(t v ; 1) because it takes a negligible time for the particles to accelerate from 0 to v (see Section III A for a further discussion). Below we analyze the settling regimes individually and explain underlying mechanisms. In Section IV we discuss how the regimes' properties and boundaries depend on the characteristics of the background ow. A. \Stone-like" regime (jv j=u & 2) t rms The simplest regime corresponds to the case with a high jv j=u ratio. For a particle with jv j > 2u , t rms t rms the average convective velocities are more than twice smaller than the speed at which the particle would be sinking if there was no convection. This implies that particles with this property are little aected by the ow { they sink almost as if the uid was at rest because thermal convection is slow relative to the particle's vertical drift. Therefore, when jv j=u & 2, Eq. (11) provides a good prediction of the settling behaviour (Fig. 3a). As t rms analyzed later, this result is very robust with respect to the values of Ra and Pr because the background ow is nearly irrelevant in this regime. For even higher jv j=u ratios, the t to Eq. (14) becomes perfect, and the acceleration from 0 to v t rms t begins to play a role in the shape of obtained solutions. In Fig. 4 we demonstrate this eect on the simulation 10 10 set xC , constructed using particles with 10 times larger radii than in the reference set C . For clarity 50 50 of the gure, we only show a few particle types from the set, with the jv j=u ratio ranging from 2 to t rms 50. As jv j and St increase in value, the analytic solution (14) loses its symmetry with respect to light and heavy particles because its second term gains in relative importance. The second term in Eq. (14) is not symmetrical with respect to v : particles with the same terminal velocity but dierent modi ed density ratio have dierent values of St (recall the de nitions of St, , and that v := St). As a result, light particles ( >1) have a shorter response time St when compared to heavy ones ( <1), and accelerate to v faster (cf. the last term in Eq. 11). We label this regime stone-like. Although particles with jv j=u & 2 still interact with the structure of t rms the ow (see Section IV A below), their vertical speed is close to the free-fall speed jv j. B. Bi-linear regime (0:3 . jv j=u . 2:0) t rms Moving to lower jv j=u ratios, the settling curves become approximately piecewise linear, with two distinct t rms settling rates. The two dierent rates correspond to dierent initial positions of the particles. In Fig. 5 we depict particles with 0:3 < jv j=u < 2:0 that are still suspended at the time t = t rms 0:4=(0:3 u ) 13. This snapshot corresponds to the time at which the settling curves of particles with rms jv j=u = 0:3 change their slope (cf. the darkest settling curve in Fig. 3b). The particles in Fig. 5 form t rms a cloud, centered above a cluster of upwellings, and the larger the value of v , the smaller is the respective cloud. The surroundings of major downwellings are free of particles. Note that we depict heavy particles only; light particles are located above the central cluster of downwellings. After t 13, all particles with 0:3 < jv j=u < 2:0 settle at a reduced rate, while up to this time their t rms settling is well captured by Stokes' law (Fig. 3b). It follows that if a heavy particle (with 0:3 < jv j=u < t rms 2:0) is initially injected close to a major downwelling, or below the top boundary layer where horizontal velocities are large, it settles quickly. Perhaps surprisingly, the settling rate of such particles is represented well by Stokes' law and does not exceed it (compare the rst linear segment of all settling curves in Fig. 3b). One could expect the downwellings to sediment the carried particles downstream, speeding up their settling beyond the rate predicted by Stokes' law. This, however, does not happen. In the rst stage of the bi-linear regime, particles touch both horizontal boundaries with little to no lateral preference (i.e. the x-coordinates Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 4. Selected settling curves from the simulation set xC that consists of particles with a high jv j=u ratio (2 t rms to 50 for the depicted selection). Solid lines show the theoretical prediction (14), while colored points are obtained from the numerical simulation. of the settling events have a uniform distribution), and the percentage of settled particles grows linearly in time, with a slope that matches the Stokes velocity. We pay further attention to the horizontal distribution of settled particles in Section IV A. In the second stage of the bi-linear regime, the settling rates are signi cantly reduced. This is because bursts of upwelling ow act against the particles' tendency to settle. The terminal velocities studied in this section are still suciently large (jv j > 0:3 u ) for the particles to eciently penetrate through the uid t rms ow, but at the same time the existence of plumes alters particle trajectories signi cantly, in particular by lifting particles that get caught in strong conduits. The higher the jv j=u ratio, the less particles survive t rms after the rst-stage settling, and the closer they are to the central axis of the major upwelling structure (compare black and white dots in Fig. 5). As a result, the settling rate is smaller for higher jv j=u ratios t rms (Fig. 3b). Note that throughout the text we refer to the relative settling rate, i.e. to the slopes of the settling curves in Fig. 3, where the x-axis represents distance rather than time. With respect to time t, the settling is generally faster for higher jv j=u ratios. t rms For jv j=u 2, the settling curves are close to being at in the second stage of the bi-linear regime. t rms This is in agreement with expectations: the velocities of plume heads are typically close to 2u and the rms carried particles thus could, in principle, be inde nitely suspended in a xed point in space by the action of a stationary plume (cf. Eq. (9) with u v ). For Ra = 10 the ow is highly non-stationary and such situation never occurs, but the idealized scenario helps explaining the very slow settling rates. C. \Dust-like" regime (jv j=u . 0:02) t rms For jv j=u . 0:3, the settling curves smoothly converge towards a single line (red line in Figs. 3c and t rms 3d). In this section we analyze this limiting case, while the transitional regime (0:02 . jv j=u . 0:3) is t rms discussed later. The settling curves are nearly identical when jv j=u . 0:02. The very existence of a limit is a non- t rms trivial result. While one can a priori expect the applicability of Stokes' law when jv j=u ! 1, when t rms jv j=u ! 0 the particles should behave as uid tracers. This can be seen directly by inspecting Eq. (9): t rms Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 5. Temperature eld in the simulation C at the time t = 12:9. The temperature range is clipped for a better visibility of the up- and downwellings. Dots show particles with 0:3 < v =u < 2:0 that have not settled by the t rms respective time. See Supplementary material, Video S1. for a xed , if the terminal velocity tends to zero, the Stokes number St also tends to zero. In the limit St ! 0, the second term on the RHS dominates Eq. (9), yielding v = u (i.e. uid tracers). For this reason, is called the response time { particles quickly adapt to the carrier uid velocity when is small. Fluid D D tracers, however, never touch the bottom nor the top boundary of the model domain and it is thus a priori unclear how particles with a small jv j=u ratio should settle. t rms A simple theoretical model for small particles (i.e. with a small Stokes number) was developed by Martin and Nokes [5]. They proposed that at the base of the model domain, where convective velocities vanish, all particles are free to settle from the uid with a speed equal to their terminal velocity v . Therefore, the rate at which the number of particles in the ow N decreases with time is given by dN v u = A v u c(0) ) N = N exp = N exp(v t); (15) t 0 0 t d H where A is the area of the base of the domain, c(z) the horizontally averaged concentration of particles at height z above the bottom boundary, and N the initial number of particles. The exponential solution in Eq. (15) is obtained by assuming c(0) to be the current average concentration N=(AH ). Indeed, Fig. 3d shows that 1 exp(v t) ts the settling curves well for jv j=u ! 0, con rming the t t rms theoretical and experimental conclusions of Martin and Nokes [5]. The derivation of Eq. (15), however, is based on a counter-intuitive assumption: particles must be uniformly distributed throughout the entire model domain by convection, and yet there must be a boundary layer with little to no mixing, thick enough for the particles to separate from the uid and accelerate to v . Moreover, the concentration of particles in the boundary layer is assumed to be the same as in the bulk of convection, without assessing the mutual transport between the two regions. In the next section, we focus on the statistics of particle transport between convection cells and boundary layers of the ow. We interpret particle settling as a random process, allowing us to provide a quantitative description of the settling regimes' boundaries, and to explain in detail why the exponential law (15) fails for particles with jv j=u & 0:02. Most importantly, our approach serves as a unifying theory, capable t rms of containing all four settling regimes in a new equation that estimates the time required for a complete sedimentation of the particles. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids D. Particle settling as a random process The velocity structure of the simulation C is shown in Fig. 6. On top of the temperature eld, shades of green show regions with juj < 0:5 u . Below, these regions are referred to as the \low-velocity piles" (the rms de nition may seem rather arbitrary here { later we investigate piles de ned generally as the regions where juj=u < p and vary the pile factor p ). The centres of the large scale convection rolls, which also show rms f f small velocities, are not considered as the low-velocity piles (thick green line in Fig. 6, explained later). Grey points in Fig. 6 show suspended particles, this time we select only those with jv j=u < 0:3. For dust-like t rms terminal velocities (jv j=u < 0:02), the particles appear uniformly distributed, while for jv j=u ! 0:3 t rms t rms the spatial distribution resembles the one from Fig. 5. In this section we explore the rate at which particles enter the low-velocity piles, and analyze how likely it is that a particle enters one but does not settle at its base and returns to fast convection cells instead. FIG. 6. Same as Fig. 5, but at time t = 129:5 and the depicted particles have smaller terminal velocities, jv j=u < t rms 0:3. In green we show regions where thermal convection is slow, with velocities below 0:5u . The thick green line rms separates the low-velocity piles from centres of large scale convection rolls, which also show small velocities. See Supplementary material, Video S2. In Fig. 7 we take a particular instant in time and compare Eq. (15) with the number of particles that have actually settled (red line vs green dots). In agreement with Fig. 3, there is a good match with the exponential law for jv j=u . 0:02, but for 0:02 . jv j=u . 1:5 the settling is slower than Eq. (15) predicts. t rms t rms Similarly as in Fig. 3, the black line represents jv jt, i.e. the prediction based on simple Stokes' settling. Note, however, that while Fig. 3 maps the temporal evolution of the settling, Fig. 7 captures only a snapshot in time and does not contain any information about the settling rate. For instance, particles with jv j=u & 1:5 t rms have completely settled at time t = 50 { both the exponential and Stokes' laws match the observation and Fig. 7 cannot be used to distinguish between the two, although Fig. 3a shows that particles with a high jv j=u ratio follow Stokes' law. t rms Blue stars in Fig. 7 mark the percentage of particles that have entered the low-velocity piles at least once by the time t = 50. For heavy particles we only consider the piles located in the bottom fourth of the model domain (thick green line in Fig. 6), for light particles only the upper fourth of the domain is considered. The center of the graph (v = 0) represents uid tracers (treated as heavy for the purpose of this analysis). It is important to notice two things: First, the probability of ever entering the low-velocity piles decreases with the jv j=u ratio, no matter how small the ratio is. This means that, even in the transitional and t rms dust-like regimes, the trajectories of inertial particles dier from those of uid tracers: particles governed by the Maxey-Riley equation are more likely to cross sluggish regions of the ow. Second, the probability of escaping the low-velocity piles and returning to the ow increases with the jv j=u ratio. This can be seen t rms Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 7. Percentage of settled particles (green dots) compared with the two analytic predictions (14) and (15) (black and red lines, respectively). Blue stars show the percentage of particles that have entered low-velocity piles at least once by the respective time. The gure corresponds to time t = 50. from the dierence between the percentage of particles that have entered the piles and the percentage of particles that have settled: while for higher values of jv j=u the two nearly coincide, for lower values they t rms dier, with the dierence growing as jv j=u approaches zero (compare blue and green symbols in Fig. 7). t rms In Fig. 8 we plot the probability of escaping the low-velocity piles. It is computed as follows: when a particle reaches a region with juj < 0:5 u , we mark it as captured. If a captured particle, instead of rms settling at the wall of the container, is transported back to a region with juj > u , we mark it as escaped. rms Each particle can be captured and escape multiple times and we store the record for each individual particle. The escape probability P is simply the sum of all escapes divided by the sum of all captures, taken over all particles of a given type. As expected, P goes to one for uid tracers, because tracers never settle and eventually always escape any low-velocity regions. On the other hand, P tends to zero for particles with a large terminal velocity, because such particles always settle upon encountering a slow region. Now we have the means to describe particle settling as a random process. The exponential law (15) and its underlying dierential equation dN = jv jN dt are, in a way, calling for such approach: in the terminology of stochastic processes, the equation says that particle settling is a Poisson counting process with intensity jv j. A random process is de ned through an event of a given probability. Here, the event is the settling to the base of a low-velocity pile, with the probability equal to 1P . For each particle type, the number of particles that settle over a time dt can be written as dN = (1P ) dN ; (16) e c where dN is the number of particles that enter the low-velocity piles in the time interval ht; t+dti. In other words, dN describes the supply of particles into our random process, and, as outlined by Fig. 7, it is a function of jv j. Under the assumption (indicated by the question mark above the equality sign) that particle settling is a Poisson counting process with intensity jv j, we may further write dN = (1P ) dN = N jv j dt: (17) e c t Eq. (17) can be validated numerically. However, due to intrinsic uctuations of thermal convection, it is convenient to integrate Eq. (17) in time and compare the integral quantities instead, 0 0 N (t ) dt (1P ) = jv j R =: jv j t : (18) e t t c dN The RHS of Eq. (18) is plotted in black in Fig. 8a. The ratio of the two integrals is labeled as t because it is the average time between particle captures, that is, between repeated entries into the low-velocity piles. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 8. (a) Settling probability 1P , where P is the probability of escaping from a region with juj=u < p , with e e rms f p = 0:5. Black symbols show v t (resp. v t for light particles), where t is the average time between repeated f t c t c c entries into the low-velocity piles. (b) Ratio of v t =(1P ) for dierent values of the pile factor p . The black dashed t c e f line shows the reference value 1 because for v t = (1P ) the solution (15) is obtained. The solid red line shows t c e the skew normal distribution f that is used in our analytic model for particle settling. By comparing 1P and jv j t , we observe that for particles with suciently small terminal velocities e t c (jv j=u . 0:02), the governing equation of the Poisson process is well satis ed (compare blue and black t rms symbols in the inset of Fig. 8a). This con rms the idea underlying the derivation in Eq. (15). Note that for a small value of p our description reduces to the one by Martin and Nokes [5]: piles with a vanishing p f f factor become only thin layers at the top and bottom boundaries, from which there is no escape (i.e. P =0). When P = 0, Eq. (17) reduces to N (t)jv j dt = dN , where the left hand side expresses the particle ux e t c that can be expected through any horizontal plane if uniformly distributed particles drift vertically with the Stokes velocity. In the limit of p ! 0, the right hand side is the supply of particles into the boundary layer. According to the assumptions of Martin and Nokes [5], these two particle uxes are equal (Eq. (15)). Our results show that this assumption is not valid generally: the value of jv j t reaches up to 4 for light t c particles in the set C (black circles in Fig. 8a). Indeed, unlike the probability 1P , the value of jv jt is e t c not limited by 1: its value merely evaluates how frequent the transport of particles between fast and slow regions of the ow is. The dierence between the ratio jv jt =(1P ) and 1 is then the observed deviation t c e from a Poisson counting process (Fig. 8b). Already for terminal velocities exceeding 0:02 u , the RHS of Eq. (18) is larger than the probability rms 1P . This explains why the exponential law (15) fails: it is too dicult for particles to enter the low- velocity piles. In other words, the transport of particles into the sluggish regions, where separation from the uid ow takes place, is much slower than what one would predict when simply assuming that the particles drift vertically with the speed jv j. Comparing jv j t and 1P provides a uni ed description of all four regimes. Dust-like: Both quantities t c e Settling of particles in thermal convection Submitted to Phys. Rev. Fluids are equal and N = N exp(jv j t) describes particle settling well. Transitional: The supply of particles into 0 t the piles is too small (jv j t > 1P ), which implies that settling is slower than that predicted by Eq. (15), t c e and the disagreement increases as jv j=u increases. Bi-linear: The supply of particles is still too small, t rms but increases quickly as jv j=u further rises, eectively reducing the dierence between jv j t and 1P . t rms t c e This is because the particles are increasingly ecient in penetrating the uid ow. Stone-like: Eventually, jv j t becomes smaller than 1P , and the settling curves become faster than exp(jv j t), soon reaching t c e t the Stokes's law instead. These newly de ned regime boundaries do not exactly overlap those used in Fig. 3, but both de nitions are in a rough agreement. While P and t depend on the value of p that is used in the de nition of the low-velocity piles e c f (i.e. juj=u < p ), their ratio does not (Fig. 8b). The quantity jv j t =(1 P ) can thus be used to rms f t c e construct a general model, extending the exponential law (15). First, we t the ratio with a skew normal distribution f : t jv j A (x ) (x ) c t = f (jv j=u ) := 0:5 + p 1 + erf p exp (19) t rms 1P 2 e 2 2 where x = jv j=u and erf is the error function. The amplitude A is prescribed as t rms 0:5 2 A := h i exp ; (20) 1 + erf in order to get the observed match between jv jt and (1P ) in the limit of zero terminal velocity (i.e. f = 1 t c e for jv j=u ! 0). The remaining parameters, ; ; and ; are obtained by tting the data (see Fig. 8b). t rms The analytic prescription (19) allows us to continue in describing particle settling as a random process. Setting dN = N dt=t in Eq. (16), we have c c (1P ) dN jv j N e t dN = (1P ) dN = N dt ) = (21) e c t dt f (jv j=u ) c t rms Eq. (21) has exponential solution: N (t) jv jt = exp : (22) N f (jv j=u ) 0 t rms Equation (22) establishes an extension to the solution (15). It is valid also for jv j=u > 0:02, and bridges t rms the gap between existing analytic solutions for small and large terminal velocities (i.e. the jv j ! 0 and jv j ! 1 limits). Note that the domain depth H must be added to the denominator when jv jt are to be t t replaced by their dimensional counterparts jv u j . Apart from Eq. (15), Martin and Nokes [5] also develop a more sophisticated theory, in which c(0) is not simply the average concentration N=AH . By assuming a depth-dependent concentration c = c(z) whose temporal changes are governed by the diusion equation [e.g. 34], Martin and Nokes [5] nd solutions for c(0)=c for several ow and particle parameters, with c being the average concentration N=AH (see their Table 2). Note that inserting c(0) 6= c into Eq. (15) is exactly analogous to our Eq. (22), with c(0)=c being analogous to our 1=f . While Martin and Nokes [5] predict c(0)=c > 1 for particles with a non-vanishing Stokes velocity, we obtain the exact opposite, f > 1. However, the experimental measurements of Martin and Nokes [5] are in agreement with our results, as they systematically measure the settling rates to be slower than Eq. (15) predicts, especially for particles with v =u 0:5. Martin and Nokes [5] acknowledge t rms the discrepancy between their theoretical prediction and measurements, and speculate that it may be related to the breakdown of the assumption of one-dimensionality. In particular, they anticipate that the large scale circulation in the uid could be responsible for the failure of the one-dimensional turbulent diusion theory, which is exactly what we observe. The mis t between Eq. (22) and the observed settling curves never exceeds 30%, with the largest error occurring for particles with 0:3 < v =u < 1:0. This is not surprising - already from Fig. 3b it is clear t rms that the settling curves are not exponential when v =u & 0:3. Nevertheless, Eq. (21) is still useful for t rms estimating the characteristic time of complete sedimentation for all particle types (see Fig. 10 at the end of this section and the accompanying discussion). Settling of particles in thermal convection Submitted to Phys. Rev. Fluids The imperfect t of the observed settling curves is caused by the fact that t is not a function of jv j only, c t but it is also a function of time, t = t (jv j; t). For instance, in the stone-like regime, the number of particles c c t N follows N (t) N (1 v t), and dN N v dt. Eq. (18) then yields t max(1=v t=2; 0:5=v ), 0 t c 0 t c t t with 0:5=v being the end value that is reached after all particles have settled (see the black line in Fig. 9, resp. the dotted line in Fig. 8b). Eq. (22), on the other hand, assumes that the average time between particle captures, t , does not vary in time (see the de nition of t in Eq. (18) versus its use in Eq. (21)). c c In Fig. 9 we show the temporal evolution of t for selected v =u ratios. Regardless of the value of c t rms v =u , t equilibrates before t 100 (note that the x-axis is logarithmic { for most of the depicted time- t rms c window t is steady). By t 100, however, most of the particles in the stone-like and bi-linear regimes have already settled (dashed lines). Re ning Eq. (22) would thus require accounting for the time dependence of t . With the black line in Fig. 9 we plot max(1=0:3t=2; 0:5=0:3), which is the theoretical value of t obtained for Stokes's settling of particles with v = 0:3 (see above). The initial, short-lived drop of the actual evolution of t (t) (dark blue line) is caused by accelerating to v from zero (the average vertical velocity of the uid, c t u dV , and thus also the initial average velocity of the particles, is zero). In the transitional and dust-like regimes, on the other hand, t (t) has only a mild temporal evolution and the transient phase is also less relevant, because the particles take longer to settle. Most of the particles with small terminal velocities settle when t (t) is completely steady, allowing Eq. (22) to provide a good t to the observed settling curves (< 10 % error for v =u < 0:1). t rms FIG. 9. Solid lines show the temporal evolution of the average time between captures, t , for selected v =u ratios. c t rms For a comparison, the characteristic time of a ow overturn is 5. The black line is equal to max(1=0:3t=2; 0:5=0:3). Dashed lines show the corresponding percentages of settled particles. In Fig. 10 the above results are summarized. In the literature, the terminal velocity v is used as a measure of the settling rate of particles in a convective ow. This implies that the time required for a complete sedimentation of all particles is 1=(v ), or H=(v u ) in terms of dimensional quantities. Fig. 10 shows the t t factor F by which 1=(v ) has to be multiplied in order to obtain the correct settling time. Since Eq. (22) is an exponential law, the predicted time of complete sedimentation is, in principle, in nite. To circumvent such inconvenience, we compute the time until 95% of particles have settled and divide the resulting value Settling of particles in thermal convection Submitted to Phys. Rev. Fluids by 0:95=v (i.e. normalize by the respective terminal time). The plotted factor, F , is therefore equal to: log(0:05) F := f (jv j=u ): (23) t rms 0:95 FIG. 10. Normalized settling time F as a function of buoyancy ratio and particle response time St. The settling time is computed with our analytic model (22), while squares, circles, and pentagons are colored according to the 10 10 10 actual settling times obtained from the numerical simulations C , xC , and B respectively. The region of relatively 50 50 50 slow settling (\slow belt" in the gure) occupies the area with 0:02 . jv j=u . 2:0. Black triangles correspond to t rms the parameters from Table I. The colour scale is clipped in order not to be dominated by the top right region, where settling occurs as if there was no convection. The settling time = terminal time ratio, F , is computed with the use of Eq. (22) when v =u < 2 and t rms from Eq. (14) when v =u > 2. We restrict the applicability of our analytic model, because for v =u > 2 t rms t rms the Stokes' formula (11) is more accurate and physically appropriate. The parameters ; ; and used in Fig. 10 come from the set C (see Table II for the respective values, Fig. 10 is plotted for heavy particles only). Solid and dashed red lines in Fig. 10 are isolines of v that correspond to the regime boundaries. Since u = 0:1, the respective values are v = 0:002, 0.04, and 0.2. For a more vigorous ow these values must rms t be adjusted accordingly (see Section V). 10 10 On top of the analytic prediction (22) we plot the settling times observed in the simulation sets B , C , 50 50 and xC (pentagons, squares, and circles respectively). Only the particle types for which the simulations have reached at least 95% settling are plotted. The dierence between Eq. (22) and the numerical simulations is typically less than 3%. One exception is the vicinity of jv j=u 0:6, where the settling rates are very t rms small during the second-stage of the bi-linear regime, which signi cantly prolongs the settling time with respect to expectations (up to 40% discrepancy in the value of F ). Nevertheless, N (t )=N 0:95 stays s 95 0 below 7% for all particle types, ensuring reasonable accuracy of Eq. (22) for general applications. Here, t is the settling time computed by setting (N N )=N = 0:95 in Eq. (22). 0 0 The region labeled \slow belt" is characterized by increased settling times, with the average settling rates being up to 13 times slower than Stokes's law predicts in the present conditions Ra = 10 and Pr = 50. With the exception of the above mentioned discrepancy between the theoretical prediction and measurements Settling of particles in thermal convection Submitted to Phys. Rev. Fluids of Martin and Nokes [5], its existence is not reported in previous literature, which calls for experimental and 3D numerical con rmation of our ndings. IV. RESULTS: IMPORTANCE OF THE BACKGROUND FLOW In the dust-like and stone-like regimes, the settling curves are robust with respect to properties of the background ow. In the transitional and bi-linear regimes, on the other hand, the above results suggest that settling curves depend on large-scale circulation of the uid (Fig. 5). Thermal convection of an isoviscous uid naturally leads to the formation of low-velocity piles similar to those depicted in Fig. 6 (see also Discussion), but their coherence and erosion depend on the values of Ra and Pr. In this section we analyze the interplay between particle settling and the velocity structure of thermal convection. A. Horizontal distribution of settled particles The escape probability P is close to zero already for jv j=u & 0:3 (Fig. 8), and the settling problem is e t rms thus mostly reduced to measuring the transport of particles from the bulk of convection into the low-velocity piles. Here we show that the near-boundary regions depicted in green in Fig. 6, i.e. the piles with p = 0:5, act as dominant sinks for most particle types. In Section III B we demonstrated that in the second stage of the bi-linear regime, heavy particles hover inside a cluster of upwellings (Fig. 5). Eventually, these particles settle in the underlying low-velocity pile, whose edges are a continuous source of the plumes that keep lifting the particles. Note that settling below upwellings is somewhat counter-intuitive { one could naturally expect heavy particles to concentrate below major downwellings. FIG. 11. Horizontal distribution of settled particles in the simulation set C . The v =u ratio is depicted in color, t rms and only the bi-linear range, 0:3 < jv j=u < 2:0, is selected. (a) Vertically averaged vertical component of the t rms uid velocity, u . The quantity is further averaged over time, with the time window of integration covering complete sedimentation of the respective particles. (b) Heavy particles. (c) Light particles. Gray symbols indicate the ow direction of two major convection cells. The settled particles are sorted into 20 equally wide bins and the plotted value is normalized by the initial number of particles, i.e. divided by N =20. Fig. 11 shows the horizontal distribution of settled particles. The distribution combines information from both stages of settling, but we note that the non-uniformity is produced in the second stage only. In the rst Settling of particles in thermal convection Submitted to Phys. Rev. Fluids stage of the bi-linear regime, the relatively fast sedimentation along with the horizontal drag associated with large-scale convection rolls ensure that particles are distributed evenly across the bottom and top boundaries (see also Video S1 in Supplementary material). For brevity, and because the horizontal distribution of light particles is analogous to the distribution of heavy particles, we will discuss only the heavy particles in this section. Top panel of Fig 11 shows the vertically- and time-averaged vertical component of the velocity eld. Peaks of the function correspond to the edges of large-scale convection cells, whose centres are indicated by the grey symbols in panels b) and c). FIG. 12. As Fig. 11, only here we show the horizontal distribution of settled particles in the simulation set xC . Fig. 12 shows that also stone-like particles settle preferentially in the large low-velocity pile located at the edges of the model domain (as long as the v =u ratio is . 10 { see the dark blue lines). Due to their t rms large sinking velocity, stone-like particles eciently cut through the bursts of upwellings. Yet, the large low-velocity pile still acts as a sink. This is because, due to the large-scale rolls, the uid velocities in the lower-half of the model domain generally point towards the side edges, which sways the sinking particles into that direction. For the particle trajectories to be close to \ballistic", i.e. to follow the analytic solution (11) exactly, the jv j=u ratios of much larger than 2 are necessary. Even when Eq. (11) eectively governs particle t rms dynamics, the horizontal distribution of settled particles is not uniform. The uid ow then enters Eq. (11) through the initial velocities, v ~(x; t=0) = u(x; t=0). Because of large-scale circulation, ballistic particles in the upper part of model domain are generally injected in the direction of major downwellings and have sucient time to move laterally. Particles injected into the lower half of the domain, on the other hand, do not have the time to move to the sides. As a result, the horizontal distribution reverses for jv j=u & 30, t rms with most of the heavy particles settling below the major downwelling (light green and yellow curves in Fig. 12). For particles with 0:02 < jv j=u < 0:3, we plot the horizontal distribution of settling in Fig. 13. In the t rms transitional regime, particles still \see" the low velocity piles, but the settling events become horizontally uniform as jv j ! 0:02. The particles experience on average over twenty ow overturns between the repeated crossings of the low-velocity piles (see Fig. 9). Typically, they circulate in large convection cells, waiting to enter the slow regions through small-scale irregularities of the ow that are produced by births of new plumes (see also the concentration of grey and black dots inside the green regions in Fig. 6). As discussed in Section IV C, for Ra = 10 the low-velocity piles are particularly coherent and do not move horizontally (see also the top panels of the gures in this section), which makes it dicult to enter them. Under such conditions, the particle supply into the piles can be particularly small and the horizontal distribution of settling events Settling of particles in thermal convection Submitted to Phys. Rev. Fluids can be highly non-uniform. FIG. 13. As Fig. 12, only here we analyze the transitional regime, 0:02 < jv j=u < 0:03. t rms In the dust-like regime, the horizontal distribution of settled particles is uniform. These, almost tracer-like particles remain suspended in the uid for very long times, repeatedly entering and leaving the low-velocity piles (Fig. 7). The escape probability P approaches 1 as jv j=u ! 0, which means that for particles with e t rms small terminal velocities the piles become transparent. Convective motions inside the piles, though relatively slow, are still fast enough to drag along the particles with a vanishing response time St (i.e. the dust-like particles). The only structure where such particles can separate from the uid becomes the thin, laterally uniform part of the no-slip boundary layer. B. Motion of light particles towards vortices ( -eect) Describing the settling behaviour with the help of jv j=u only is a crucial reduction of the ve- t rms dimensional model parameter space. Given the complexity of the problem at hand, such description can only be used as a rst-order approximation. One diculty appears already in Fig. 8a, since there is clearly an asymmetry between light and heavy particles that have the same amplitude of the terminal velocity. For Ra = 10 , the dierences are small and may partially result from a random asymmetry of the up-and downwelling regions. For Ra = 10 , the asymmetry becomes a prominent feature (Fig. 14). Its underlying mechanism is explained below. Due to the term Du=Dt in Eq. (9), heavy particles ( <1) have a tendency to move away from strong ow vortices [57]. Indeed, in Fig. 15, where we show a snapshot from the simulation C , there is a reduced concentration of heavy particles at the edges of the particle cloud, i.e. close to the centres of the two largest convection rolls (in Fig. 5 this eect is not observed because the Reynolds number is too small). Light particles, on the other hand, move toward ow vortices thanks to the Du=Dt term [16]. In Fig. 16, we show only the particles with v =u 0:3 (the light/heavy asymmetry seems largest for this value, t rms see Fig. 14). Around strong and long-lived vortices, there is an increased concentration of light particles. These particles are trapped until the respective vortices vanish, which explains the enhanced attening of the settling curves. Due to the way particle sets B and C are constructed, the range of is approximately h0:99; 1:01i for one and h0:6; 3:0i for the other. Since the outward (resp. inward) motion of heavy (resp. light) particles depends on how much departs from unity, the eects encountered in Figs 15 and 16 for the set C do not occur 50 Settling of particles in thermal convection Submitted to Phys. Rev. Fluids 12 12 12 FIG. 14. Same as Fig. 3b, only here for the simulation sets B (left) and C (right). While B contains particle 50 50 50 types with the modi ed density ratio very close to 1 and the resulting settling curves show a strong light/heavy symmetry, for the simulation C the situation is dierent. Due to the -eect, heavy particles are pushed away from ow vortices and light particles towards them. This results in faster settling for heavy particles and an extreme attening of the settling curves of light particles. FIG. 15. Similar to Figs. 5 and 6, only here for the simulation set with the highest Rayleigh number: C . The temperature scale is clipped, with the respective range being depicted in Fig. 16 below. The snapshot is taken at t = 0:4=(0:3 u ), i.e. at the time for which the terminal distance is 0.4 for the settling curve with v =u = 0:3 (as rms t rms in Fig. 5). See Supplementary material, Video S3. 12 12 12 in B . In terms of the settling curves, B shows a symmetry between light and heavy particles, while C 50 50 50 exhibits a slightly enhanced settling of heavy particles and a signi cantly delayed settling of light particles (Fig. 14). Each simulation set contains several particle types with roughly the same jv j=u ratio. For C and t rms v =u = 0:3 0:02, the modi ed density ratio ranges from 1.4 to 2.5 (see Fig. 1). The respective t rms particle types are shown in Fig. 16. For these few particle types, the concentration in the vicinity of stable vortices seems similar, and also their settling curves are comparable. Based on this particular example, we can crudely conclude that for 2 h1:4; 2:5i the -eect is similarly strong and for < 1:01 no eect is observed. A more detailed analysis should be the subject of future work. One way to quantify the -eect as a function of v =u is through the v t =(1P ) ratio. Already in t rms t c e Fig. 8 there was a dierence between light and heavy particles, and the gap further increases as the Reynolds 10 10 12 number increases. In Fig. 17 we plot v t =(1P ) for the sets B , C , and C . Consistently with the t c e 10 10 50 Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 16. Demonstration of the -eect. When convective vigor is high, particles from the transitional and bi-linear range, 0:02 < jv j=u < 2:0, have a tendency to move away, resp. towards strong ow vortices. Here we show the t rms increased concentration of light particles with v =u = 0:3 0:02 near long-lived vortices that have developed in t rms the simulation set C . See Supplementary material, Video S4. 10 10 12 FIG. 17. jv jt =(1P ) ratio for the simulation sets B , C , and C . Light particles (circles) experience slower t c e 10 10 50 settling than the heavy ones, as long as the -eect comes into play (sets labeled as C). The separation between the ratios for light and heavy particles increases with the Rayleigh number (cf. the orange and green symbols). analysis above, light particles settle very slowly in C , with v t =(1P ) going up to 14 for particle types t c e that are trapped inside vortices for a particularly long time, while there is little dierence between the settling rates of light and heavy particles in the set B . Note that for heavy particles the value of v t =(1P ) may drop below 1 due to the -eect (green triangles t c e in Fig. 17). This result can be interesting in view of the debate between Wang and Maxey [18] and Mei [58]: the former observed faster than Stokes' settling of heavy particles due to preferential sweeping in downward moving uid, while the latter did not observe the eect [see also 19, 59]. Here we nd an increased settling rate that is higher than predicted by the exponential law of Martin and Nokes [5], but it never exceeds the Settling of particles in thermal convection Submitted to Phys. Rev. Fluids Stokes' velocity on average. Note, however, that our Pr is always relatively high, limiting turbulence eects. C. Eects of convective vigor and uid inertia In this section we analyze how the above results are aected by Ra and Pr of the background ow. In Table IIa we show the parameters ; ; and of the skew normal distribution f for all the simulation sets labeled as B, along with the maximum value of f , denoted as m, and the value of jv j=u at which the t rms maximum is reached. The maximum m can be used as an estimate for how much slower the particle settling can be when compared to the exponential decay N = N exp(jv jt). 0 t TABLE II. a) Parameters of distribution f , no -eect. Set Reynolds n. max. m at jv j=u t rms B 100 0.04 0.64 3.46 1.5 0.32 B 430 0.18 0.51 2.25 2.4 0.45 B 1440 0.18 0.43 2.43 3.2 0.39 B 6770 0.21 0.44 1.84 2.6 0.45 B 21790 0.23 0.40 0.81 1.6 0.42 b) Parameters of distribution f , with -eect. Simulation set max. m at jv j=u t rms Heavy, C 0.02 0.82 7.44 1.6 0.25 Light, C 0.04 0.69 4.73 1.7 0.31 Heavy, C 0.11 0.60 1.93 1.6 0.44 Light, C 0.25 0.39 1.77 4.0 0.46 Heavy, C 0.14 0.60 4.10 3.3 0.39 Light, C 0.20 0.51 3.09 4.3 0.44 Heavy, C 0.11 0.60 1.94 1.6 0.44 Light, C 0.25 0.39 1.77 4.0 0.46 Heavy, C 0.87 1.01 -3.50 1.1 0.42 Light, C 0.28 0.35 2.01 9.6 0.47 The most important outcome of the comparison in Table IIa is that the critical ratios jv j=u that mark t rms the regime boundaries are largely independent of Ra and Pr: the maximum of the function f , i.e. the boundary between the transitional and bi-linear regimes, always lies at jv j=u 0:4 0:1. With a similar t rms accuracy, the stone-like regime is always obtained for jv j=u 2:0 (recall that Stokes' settling satis es t rms v t =(1P ) = 0:5, see the dotted line in Fig. 8b). t c e The settling behaviour in the limits jv j=u ! 1 andjv j=u ! 0 can be derived analytically regardless t rms t rms of the values of Ra and Pr: in the rst case, the ow is irrelevant. In the second case, the only requirements are those discussed above in relation to Eq. (15). The critical values of jv j=u for which the settling t rms curves start to substantially deviate from these limits, could, however, strongly depend on the Rayleigh and Prandtl numbers. The fact that there seems to be no such dependence makes the possibility to extrapolate our results to arbitrary thermal ows promising. The only notable dierence between the various simulation sets is the width of the bi-linear and transitional bands, i.e. the spread of the settling curves in the transitional and bi-linear regimes. The width of the band 12 12 is directly linked with the maximum value m (compare e.g. the sets B and C in Figs 14 and 17), and m 50 50 depends non-trivially on Ra and Pr. Generally, there is a trend between m and the Reynolds number, with 3 4 m being the largest for Re 2 [10 ; 10 ]. Increasing the Reynolds number Re enhances the short-wavelength content of the velocity eld, and alters the stability of large-wavelength structures. In particular, the low-velocity regions become less stable (see Fig. 15). High convective vigor is capable of tearing the sluggish, boundary-based piles into chunks that are advected into the rest of the uid and mixed. This results in spatial variations of large-scale circulation and thus in faster settling because the sinking particles spread across a broader area (see Videos S2 and S4 in Supplementary material). Settling of particles in thermal convection Submitted to Phys. Rev. Fluids Surprisingly, particles with 0:02 < jv j=u < 2:0 show faster settling also when Ra, resp. Re decreases. t rms As expected, in the simulation C the plumes are thicker and live longer than those in the simulations 10 12 C and C . The thicker and well separated plumes, however, allow particles to sink in between them, 50 50 which results in an increased settling rate in the second stage of the bi-linear regime when compared to the reference set C (see also Video S5 in Supplementary material). For sets labeled as C and Re 10 , the -eect splits the v t =(1P ) ratio into two clearly distinct t c e functions. In Table IIb we provide the corresponding sets of parameters ; ; and . The dierence between m for the heavy and light particles increases with the Reynolds number. V. APPLICATION TO CRYSTALLIZING MAGMA 9 17 The Rayleigh number of a magma chamber is of the order of 10 {10 [e.g. 60], making our results directly applicable to relatively small volumes of magma. For a global magma ocean, however, Ra 10 (Table I). First of all, in order to apply our results to this system, it is necessary to estimate u for such an extreme rms ow regime. In terms of non-dimensional control parameters, the v =u ratio can be expressed as: t rms v St u St gTH H St Ra = = = p ; (24) u U Re Re Pr rms rms where the Reynolds number is de ned through the volume-averaged root mean square velocity, Re := U H=. rms Upon employing a Re = Re(Ra; Pr) scaling, Eq. (24) can be used to compute the jv j=u ratio for various t rms Rayleigh and Prandtl numbers. Relationships between Reynolds number and Rayleigh and Prandtl numbers have been obtained based on various experimental, numerical, and theoretical work. Here we adopt the Grossmann-Lohse theory [61]. The theory de nes four regimes for isoviscous thermal convection, depending on whether kinetic and thermal energy dissipation takes place dominantly in the boundary layer region, or in the convective bulk. For the ranges of Ra and Pr investigated here, the energy dissipation is dominated by the convective bulk, and the thermal boundary layer is nested inside the kinetic one. In an idealized case such situation 4=9 2=3 yields Re / Ra Pr (see Table 2, regime IV in Ahlers et al. [37]). Based on our 5 data points, we 4=9 2=3 (1:310:02) observe Re = (0:070:01) (Ra Pr ) , i.e. our exponents dier by ca. 30% from the idealized case. However, the Grossmann-Lohse theory is derived for a 3D box with thermally insulating side-walls, while we perform 2D simulations with periodic sides, which may be a source of the discrepancy (see also Discussion). When extrapolating to Ra = 10 according to our Re(Ra; Pr) relationship, Eq. (24) gives u [5:2; 2:9] for Pr = [10; 50]. rms In Fig. 18a we show the settling time of crystals in a global magma ocean (Table I). The results are obtained by multiplying the terminal time t := 1=jv j (resp. H=(v u ) in dimensional units) with the factor t t t F = F (jv j=u ). The black triangles in Fig. 18a correspond to heavy particles in the particle set A, t rms i.e. to the black triangles from Fig. 10. The mean velocity of the ow is assumed to be u = 4:0, a value rms representative of the relevant range [2:9; 5:2]. Since u is now larger than in Fig. 10, the slow belt moves to rms the right with respect to the positions of the particle types, which now fall into the dust-like and transitional ranges (see the red lines and black symbols in Figs 10 and 18a). The factor F is computed using the a coecients derived for the particle set C , as this simulation has i 50 the highest Ra and includes the -eect. F reaches a maximum of 19 for light particles, while for the heavy ones it only slightly exceeds 3 (see the light/heavy asymmetry in Fig. 18a). Note that there are three independent parameters in the simpli ed Maxey-Riley equation (9), but in the dimensional version (8) there are only two, and , because the gravitational acceleration g is usually xed. Considering St ; and as independent is thus a generalization for arbitrary gravity. In Fig. 18 gravity is xed again, g = 9:8 m/s , and the two independent variables are chosen as = and r , because p f c these quantities are typically measured. The settling times that we obtained tend to be small compared to the typical lifetimes of magma oceans [e.g. 32, 33]. This suggests that a magma ocean will likely solidify via fractional crystallization. An exception is for particle types with extremely small density contrasts. For example, for r = 1 mm, the value of j = 1j c p f must be smaller than 4 10 in order to obtain settling time longer than 1 Myr, while the typical values are ca. [0.01,0.25] in a cooling magma [e.g. 6]. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids FIG. 18. (a) Settling time of crystals with r 2 [0:1; 100] mm and = 2 [0; 2], suspended in a global magma ocean c p f with parameters given in Table I, for which we estimate u to be [2:9; 5:2]. The black triangles correspond to the rms black triangles in Fig. 10, circles are the respective light particle types. We show log of the value of t F , where t is t t the terminal time H=(v u ) and F accounts for the characteristic settling behaviour of the various particle types (see Eq. (23) and Table II, the parameters ; ; and correspond to the set C , i.e. to the simulation with the highest Ra). (b) Same as above, only this time u = 0:5 instead of 4.0 and the size of the system is H = 2:9 km instead rms of H = 2890 km. The particle set now spreads across the slow belt. We show only the heavy particles from the set (the position of particle types and the red lines are symmetrical with respect to 1 on the x-axis). While the position of the slow belt in the (St; )-diagram depends on the value of u only, the positions rms of crystals of given sizes and radii in that diagram depend on various other parameters (see Eq. (10)). Similarly as for a global magma ocean, we can provide rst-order estimates also for magma chambers. The key dierence between the two is in the value of H . When H = 3 km instead of H 3 10 km, crystals ref with the same radius range r = [0:5; 10] mm move by 3 orders of magnitude to the right in Fig. 10. At the same time, the lower value of H reduces the Rayleigh number to 10 , yielding u 0:5, which rms shifts the slow belt by a factor of 5 to the right in Fig. 10. As a result, the crystals span over all the settling regimes, with 0.5 mm crystals being in the dust-like regime for = up to 1.5, and 10 mm crystals being p f in the stone-like regime for = larger than 1.1 (Fig. 18b). Again, the expected density contrasts result in p f settling times signi cantly smaller than the life-span of the system [typically more than 100 kyr, see e.g. the review 62], indicating fractional crystallization. The position of the studied particles with respect to the slow belt has important consequences, as it is directly related to the horizontal distribution upon sedimentation. Thus, for a mantle-deep magma ocean, the crystal radius must be & 10 mm in order to experience a horizontally non-uniform accumulation of sediments (see the black triangles that fall into the transitional regime in Fig. 18a). For a 3-km-deep magma chamber, on the other hand, the overlap between [0:5; 10] mm crystals and the slow belt indicates that the majority of suspended particles will eventually settle in the low-velocity piles (under the assumption that a large scale circulation is present and the low-velocity piles form, see also Discussion). Note that the rhythmic sedimentation suggested by Sparks et al. [38] intrinsically relies upon the assump- tion that precipitated crystals settle much faster in a non-convecting uid than when convection is present. The existence of the slow-belt presented here is thus in favour of the scenario proposed by Sparks et al. [38], for which the authors found petrological evidence in fully solidi ed chambers. VI. DISCUSSION The crystallization of a primordial molten mantle is a complex system in which the generation, settling, and re-entrainment of crystals are competing processes. Here we only focused on one of these components: the settling of crystals. Typical time scales for the solidi cation of a whole-mantle terrestrial magma ocean range from 10 Settling of particles in thermal convection Submitted to Phys. Rev. Fluids years in the absence of atmosphere, up to 10 years in the presence thereof [e.g. 32, 33]. When compared to these time scales, our results indicate a fast settling (Fig. 18a), and thus support the idea of a fully fractional crystallization. In the series of papers by Solomatov [for a review, see 8], it is instead argued for equilibrium crystallization of the majority of the primordial mantle. This is largely because re-entrainment of sedimented particles from the bottom of the uid is claimed to be the dominant process. Solomatov et al. 2 2 [63] derive a formula for the equilibrium crystal fraction, = 18Q=(gc r ), where Q is the surface eq p heat ux and denotes the fraction of available convective energy that goes into re-entrainment, estimated to be 0:1 1 % [63]. For the simulations presented here, e.g. for Ra = 10 ; Pr = 50 and crystals with = = 0:1 and r = 1 cm, the resulting is only around 3 % (the formula for is designed for a single c eq eq type of crystals only; for a range of crystal properties it must be decided how much of the available energy goes into the lifting of the various types). For the signi cantly larger heat uxes that accompany the early stages of a global magma ocean solidi cation, it quickly reaches 100%, indicating full suspension [64]. More recently, the scaling law of Solomatov et al. [63] was con rmed by the experimental study of Lavorel and Le Bars [19], who systematically varied the density ratio = and the temperature contrast that drives p f thermal convection. An important nding of their study is that in a highly turbulent ow the molecular viscosity that appears in the formula for the Stokes' velocity must be replaced by an apparent viscosity in order to account for turbulent eddies smaller than the particles. Such approach was successfully used to describe the dynamics of nite-size particles in turbulent ows [65, 66], but it is dicult to generalize in non-homogeneous ows such as for thermal convection between parallel plates. In the context of particle settling, Lavorel and Le Bars [19] found this approach viable, and the apparent viscosity that they measured at Ra = 3 10 was only ca. 2.7 times larger than the molecular viscosity. We note, however, that replacing molecular viscosity with apparent (or turbulent) viscosity in their formula for the decay of the number of suspended particles (their Eq. (9)) is similar to dividing the terminal velocity jv j by f as in our Eq. (22), resp. it is mathematically identical if f can be treated as constant for the investigated particles. Moreover, the amplitude 2.7 is within the range that we obtain for f . In other words, the decrease of the Stokes' velocity that Lavorel and Le Bars [19] computed to reconcile their measurements may have been caused by the eects of large-scale circulation analyzed in this paper as well as by the eect of sub-particle sized turbulence. Similarly to the previous works, Lavorel and Le Bars [19] used a cuboid tank lled with salty water and spherical, polymethyl methylacrylate particles. We note that the conditions of the experiments [5, 63, 67] on which the energetic analysis of Solomatov and Stevenson [36] is based dier signi cantly from the environment of non-spherical silicate crystals that accumulate at the bottom of a cooling magma. In nature, the sedimented crystals may be subject to chemical and petrological altering, possibly binding the crystals together, i.e. making them prone to re-entrainment. Note that fractional crystallization is often reported in exposed plutons [e.g. 38]. While we do not argue against re-entrainment as such, we merely point out that its workings should be thoroughly investigated in future work in the context of magma environments. In fact, the non-uniform horizontal distribution of settling events that we observe in the transitional and bi-linear regimes is slightly in favour of re-entrainment. As discussed in Solomatov et al. [63], embedded particles may be lifted by the tangential stresses caused by rising plumes, and the respective stresses increase with distance from the domain boundary. We observe a large concentration of sedimented particles in the low-velocity piles. Inside these regions, the sediments would thus build tall dunes. Since most of young plumes are born at the edges of the low-velocity piles, the crests of these dunes should be exposed to large tangential stresses. In terms of the predicted settling time, some of our results dier dramatically from those of Verhoeven and Schmalzl [40]. Roughly speaking, for particles satisfying ( ) < T (see their Fig. 12 and p f f Eq. 5), they obtain a temperature-dominated convection mode (T-regime), in which the ow is thermally driven and all particles are held inde nitely in suspension (similar results are obtained in the non-rotating cases of Maas and Hansen [43, 44]). This is because in their formulation the momentum equation is solved for the volumetric average of the uid and particle velocity, i.e. their particles by de nition follow the uid as described in Eq. 6 in Verhoeven and Schmalzl [40] (but they can also invoke uid motion, see the next paragraph). Therefore, in case of vigorous thermal mixing of the uid, their particles never settle. In our study, particles can have dierent velocities from the surrounding uid, which allows uid-particle separation and thus sedimentation regardless of convective vigor. For stronger density contrasts, roughly for ( ) > T , Verhoeven and Schmalzl [40] obtain a p f f particle-driven convection in which a layer of sediment is segregated from the rest of the uid (C-regime). We note that the condition ( ) > T simply means that the critical concentration that is required p f f Settling of particles in thermal convection Submitted to Phys. Rev. Fluids for the formation of a settling front, as described earlier by Koyaguchi et al. [6] and Sparks et al. [38], is less than 100%. In other words, the C-regime is established whenever the formation of a settling front (and thus cessation of convection due to particle motion) can take place for some critical particle concentration C < 100 %. As analyzed by Solomatov and Stevenson [36], for the crystals of interest in magma oceans and chambers, C is typically less than 100% (and this is also the case for most of the particle types investigated in the present paper). Within the C-regime, Verhoeven and Schmalzl [40] develop a model that is based on the theory of Martin and Nokes [5]. In particular, they complement the theory by accounting for the volume occupied by particles that have already sedimented. This is not to be confused with our model, in which the function f is a measure of the rate of particle transport into the low-velocity piles, normalized by the probability of not escaping from these regions. As such, our function f depends non-trivially on the structure of the background ow, while the factor f in Eqs. 26{32 of Verhoeven and Schmalzl [40] represents packing of sedimented material. Verhoeven and Schmalzl [40] then verify their model on a set of simulations in which the employed particles fall into the dust-like regime according to our classi cation (see their Eq. 8, Fig. 10, and the parameters listed below Fig. 15). This should be understood as yet another con rmation of the applicability of Eq. (15) for particles with a small value of the jv j=u ratio. t rms At the beginning of our simulations, we inject the particles uniformly throughout the entire model domain. In systems where the particulate phase is a product of a chemical reaction or phase change, this is typically not the case. In particular, in a cooling magma, the solid crystals nucleate in the relatively cold downwellings and degassing takes place due to decompression in hot rising plumes. The importance of the initial positions of newly formed particles depends on their v =u ratio. On one hand, in the dust-like regime the particles t rms are likely to get thoroughly mixed and the settling rates would not be aected. In the stone-like regime, on the other hand, the particles' trajectories and settling times strongly depend on the particles' initial positions and velocities. Given the particle positions associated with the second stage of the bi-linear regime (Fig. 5), faster settling than reported here is to be expected on average if the majority of heavy crystals form preferentially in downwellings. For Ra = 10 and Pr = 50, i.e. for the highest investigated Reynolds number, the Kolmogorov length 3 3 scale is 210 , while the upper bound of the non-dimensional particle radius r =H is 1:510 (i.e. the smallest turbulent eddies are only slightly larger than the investigated particle radii in the respective simulation set). For the simulation set xC the assumed crystal sizes even exceed the Kolmogorov length, and the particle Reynolds numbers are in turbulent rather than laminar regime (i.e. breaching the range of validity of Eq. (9)). We note, however, that the set xC is performed only for illustrative purposes in the context of explaining some aspects of the stone-like regime. Our study shows that the existence of a stable large-scale ow structure has a clear signature on the settling of particles. First, it delays the settling on average. Second, it is responsible for the non-uniform horizontal distribution of settling events. Since the large-scale circulation is observed at the highest Rayleigh numbers reached so far in numerical simulations [68] and experiments [37], we speculate that this can be an important feature for the extreme regime of a cooling magma ocean. Superstructures were analyzed in detail in 3D systems with large aspect ratios by Pandey et al. [69]. To test the in uence of a larger aspect ratio, we performed the reference simulation C also with aspect ratio 4, and the resulting function f as well as the amplitude of the horizontal variations of settling events were nearly identical. 27 0:582 0:873 In Section V we performed an extrapolation to Ra = 10 , assuming Re = 0:07 Ra Pr . Gross- mann and Lohse [61] distinguish four idealized regimes of convection depending on whether kinetic and thermal dissipation rates are dominated by the convective bulk or the boundary layers. For our Rayleigh 4=9 2=3 and Prandtl numbers, the idealized case (\pure power-law") yields Re / Ra Pr , but already for 1=2 1=2 Pr 1 we would be on the boundary with the Re / Ra Pr regime, in which the kinetic energy dissipates in the velocity boundary layer of the ow. While these are idealized cases, real convection is a mixture of these regimes, and for a detailed treatment one must employ the full theory of Grossmann and Lohse [61]. Using scaling laws to extrapolate to very high Rayleigh numbers, such as done in Section V, is, however, still subject to an open debate [e.g. 37]. It is interesting to note that the amplitude of u does not depend on Ra and Pr in the limit of Re / rms 1=2 1=2 Ra Pr , i.e. in the idealized scenario for low Prandtl and high Rayleigh numbers. In such a case, the settling behaviour could be predicted simply by computing the terminal velocities of particles of interest since the v =u ratio would not depend on the exact values of Ra and Pr (see Eq. (24)). However, for low t rms values of Pr new settling regimes may exist, which we plan to investigate in the future. Convection in magma oceans occurs in the presence of rotation, which is not included in our model. Using Settling of particles in thermal convection Submitted to Phys. Rev. Fluids estimated values for Earth's magma ocean, the convective Rossby number lies in a range of 0.03 - 100 [43], and the large-scale circulation thus may be disrupted due to rotation [70, 71]. Both the settling rates and the horizontal distribution of sedimented material would be aected in such scenario. Recently, the eects of rotation on the distribution of crystals and the rate of settling in the early stages of Earth's primordial magma ocean were analyzed in spherical geometry by Maas and Hansen [44]. In a rotation dominated scenario (Rossby . 0:3) they nd a much pronounced settling as convection and thus vertical entrainment of particles is suppressed. VII. SUMMARY We evaluate the settling rate of inertial particles that are injected into statistically steady state thermal convection. Previously, the number of suspended particles in such system was assumed to follow either the relation N = N exp(v t) or N = N (1 v ), with v being the Stokes' velocity. We observe a new 0 t 0 t t regime with particularly slow sedimentation, in which large-scale circulation prevents particles from reaching the boundary layers of the uid. By introducing a new framework that treats the settling mechanism as a random process, we develop a model that uni es the observed settling rates into the general equation N = N exp(v t=f ), where f is a function of the ratio of Stokes' and mean characteristic velocity of the 0 t ow, v =u . We investigate f over a broad range of Reynolds numbers and show that the function is t rms relatively robust. It reaches its maximum for v =u 0:4, the maximum value ranging approximately t rms from 1.5 to 3 for particles with mild density contrasts with respect to density of the uid (Table IIa). We also analyze the horizontal distribution of settled particles. Within the regime of slow settling, heavy particles accumulate preferentially below major clusters of upwellings. These are located at edges of large- scale convection rolls. For Reynolds numbers larger than 1000 and particles with a stronger density contrast, additional complexity arises because of the preferential concentration phenomenon, i.e. light particles have a tendency to move towards ow vortices, while heavy particles move away from them. As a result, light particles get captured inside long-lived vortices, which signi cantly prolongs their sedimentation at the top boundary. The maximum value of f is close to 10 (resp. the normalized settling time F 30) for our simulation with the highest Rayleigh number, Ra = 10 . When extrapolated to the extreme conditions of solidifying magma chambers and oceans, our results pre- dict fractional crystallization. For a better understanding of such complex systems, it is possible (and nec- essary) to extend our method to account for 3D geometry, rotation, re-entrainment of sedimented particles, self-consistent nucleation of solid crystals, and the coupling between particle concentration and momentum conservation of the uid. VIII. ACKNOWLEDGEMENT We thank Tina Ruc kriemen for useful discussions. VP and NT acknowledge support from the Helmholtz Association (project VH-NG-1017) and the DLR-DAAD research fellowship program. [1] H. F. Schwaiger, R. P. Denlinger, L. G. Mastin, Ash3d: A nite-volume, conservative numerical model for ash transport and tephra deposition, J. Geophys. Res. 117 (2012). [2] N. Helbig, B. Vogel, H. Vogel, F. Fiedler, Numerical modelling of pollen dispersion on the regional scale, Aerobiologia 20 (2004) 3{19. [3] J. Ruiz, D. Macias, F. Peters, Turbulence increases the average settling velocity of phytoplankton cells, Proc. Natl. Acad. Sci. USA 101 (2004) 17720{17724. [4] K. Squires, H. Yamazaki, Preferential concentration of marine particles in isotropic turbulence, Deep Sea Res. Oceanogr. Res. Pap. 42 (1995) 1989{2004. [5] D. Martin, R. Nokes, A uid-dynamic study of crystal settling in convecting magmas, Jour. Petro. 30 (1989) 1471{1500. [6] T. Koyaguchi, M. A. Hallworth, H. E. Huppert, R. Stephen J. Sparks, Sedimentation of particles from a convecting uid, Nature 343 (1990) 447{450. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids [7] L. T. Elkins-Tanton, Magma Oceans in the Inner Solar System, Ann. Rev. Earth Pl. Sc. 40 (2012) 113{139. [8] V. Solomatov, 9.04 - Magma Oceans and Primordial Mantle Dierentiation, in: G. Schubert (Ed.), Treatise on Geophysics (Second Edition), Elsevier, Oxford, second edition edition, 2015, pp. 81 { 104. [9] D. Breuer, T. Rueckriemen, T. Spohn, Iron snow, crystal oats, and inner-core growth: modes of core so- lidi cation and implications for dynamos in terrestrial planets and moons, Prog. Earth Planet. Sci. 2 (2015) [10] C. A. Jones, Thermal and Compositional Convection in the Outer Core, in: G. Schubert (Ed.), Treatise on Geophysics (Second Edition), Elsevier, Oxford, second edition edition, 2015, pp. 115 { 159. [11] L. Zhang, X. Lv, A. T. Torgerson, M. Long, Removal of impurity elements from molten aluminum: A review, Mineral Processing and Extractive Metallurgy Review 32 (2011) 150{228. [12] B. H. Chang, A. F. Mills, E. Hernandez, Natural convection of microparticle suspensions in thin enclosures, Int. J. Heat Mass Transfer 51 (2008) 1332{1341. [13] F. Toschi, E. Bodenschatz, Lagrangian Properties of Particles in Turbulence, Annu. Rev. Fluid Mech. 41 (2009) 375{404. [14] G. A. Voth, A. Soldati, Anisotropic Particles in Turbulence, in: Davis, SH and Moin, P (Ed.), Annu. Rev. Fluid Mech., volume 49 of Annual Review of Fluid Mechanics, 2017, pp. 249{276. [15] V. Mathai, D. Lohse, C. Sun, Bubbly and buoyant particle{laden turbulent ows, Annual Review of Condensed Matter Physics 11 (2020) 529{559. [16] M. Maxey, The motion of small spherical particles in a cellular ow eld, Phys. Fluids 30 (1987) 1915{1928. [17] C. Pasquero, A. Provenzale, E. Spiegel, Suspension and fall of heavy particles in random two-dimensional ow, Phys. Rev. Lett. 91 (2003). [18] L. Wang, M. Maxey, Settling velocity and concentration distribution of heavy-particles in homogeneous isotropic turbulence, J. Fluid Mech. 256 (1993) 27{68. [19] G. Lavorel, M. Le Bars, Sedimentation of particles in a vigorously convecting uid, Phys. Rev. E 80 (2009). [20] V. S. Safronov, Evolution of the Protoplanetary Cloud and Formation of the Earth and the Planets, NASA, 1969. Israel program for scienti c translations. [21] W. B. Tonks, H. J. Melosh, Core formation by giant impacts, Icarus 100 (1992) 326{346. [22] W. B. Tonks, H. J. Melosh, Magma ocean formation due to giant impacts, J. Geophys. Res.: Planets 98 (1993) 5319{5333. [23] N. Tosi, S. Padovan, Mercury, Moon, Mars: Surface expressions of mantle convection and interior evolution of stagnant-lid bodies, in: H. Marquardt, M. Ballmer, S. Cottar, K. Jasper (Eds.), Mantle convection and surface expressions, AGU Monograph Series, Wiley, 2020, pp. 1 { 45. arXiv:1912.05207. In press. [24] N. Tosi, A. C. Plesa, D. Breuer, Overturn and evolution of a crystallized magma ocean: A numerical parameter study for Mars, J. Geophys. Res. Planets 118 (2013) 1512{1528. [25] A.-C. Plesa, N. Tosi, D. Breuer, Can a fractionally crystallized magma ocean explain the thermo-chemical evolution of Mars?, Earth Planet. Sci. Lett. 403 (2014) 225{235. [26] M. D. Ballmer, D. L. Louren co, K. Hirose, R. Caracas, R. Nomura, Reconciling magma-ocean crystallization models with the present-day structure of the Earth's mantle, Geochem. Geophys. Geosyst. 18 (2017) 2785{2806. [27] M. Maurice, N. Tosi, H. Samuel, A.-C. Plesa, C. Huttig, D. Breuer, Onset of solid-state mantle convection and mixing during magma ocean solidi cation, J. Geophys. Res. Planets 122 (2017) 577{598. [28] B. Marsh, M. Maxey, On the distribution and separation of crystals in convecting magma, J. Vol. Geo. Res. 24 (1985) 95{150. [29] D. Martin, R. Nokes, Crystal settling in a vigorously convecting magma chamber, Nature 332 (1988) 534{536. [30] R. Sparks, Dynamics of magma degassing, Geological Society, London, Special Publications 213 (2003) 5{22. [31] L. T. Elkins-Tanton, Linked magma ocean solidi cation and atmospheric growth for Earth and Mars, Earth Planet. Sci. Lett. 271 (2008) 181{191. [32] T. Lebrun, H. Massol, E. Chasse ere, A. Davaille, E. Marcq, P. Sarda, F. Leblanc, G. Brandeis, Thermal evolution of an early magma ocean in interaction with the atmosphere, J. Geophys. Res. Planets 118 (2013) 1155{1176. [33] A. Nikolaou, N. Katyal, N. Tosi, M. Godolt, J. L. Grenfell, H. Rauer, What Factors Aect the Duration and Outgassing of the Terrestrial Magma Ocean?, Astrophys. J. 875 (2019). [34] R. W. Bartlett, Magma convection, temperature distribution, and dierentiation, American Journal of Science 267 (1969) 1067{1082. [35] H. Huppert, R. Sparks, The uid-dynamics of a basaltic magma chamber replenished by in ux of hot, dense ultrabasic magma, Cont. Min. Petro 75 (1980) 279{289. [36] V. Solomatov, D. Stevenson, Suspension in convective layers and style of dierentiation of a terrestrial magma ocean, J. Geophys. Res. 98 (1993) 5375{5390. [37] G. Ahlers, S. Grossmann, D. Lohse, Heat transfer and large scale dynamics in turbulent rayleigh-b enard con- vection, Rev. Mod. Phys. 81 (2009) 503{537. [38] R. S. Sparks, H. E. Huppert, T. Koyaguchi, M. A. Hallworth, Origin of modal and rhythmic igneous layering by sedimentation in a convecting magma chamber, Nature 361 (1993) 246{249. Settling of particles in thermal convection Submitted to Phys. Rev. Fluids [39] T. H oink, J. Schmalzl, U. Hansen, Dynamics of metal-silicate separation in a terrestrial magma ocean, Geo- chemistry, Geophysics, Geosystems 7 (2006). [40] J. Verhoeven, J. Schmalzl, A numerical method for investigating crystal settling in convecting magma chambers, Geochem. Geophys. Geosys. 10 (2009). [41] P. Cundall, O. Strack, Discrete numerical model for granular assemblies, Geotechnique 29 (1979) 47{65. [42] H. J. Park, K. O'Keefe, D. H. Richter, Rayleigh-Benard turbulence modi ed by two-way coupled inertial, nonisothermal particles, Phys. Rev. Fluids 3 (2018). [43] C. Maas, U. Hansen, Eects of earth's rotation on the early dierentiation of a terrestrial magma ocean, Journal of Geophysical Research: Solid Earth 120 (2015) 7508{7525. [44] C. Maas, U. Hansen, Dynamics of a terrestrial magma ocean under planetary rotation: A study in spherical geometry, Earth and Planetary Science Letters 513 (2019) 81 { 94. [45] V. Mathai, E. Calzavarini, J. Brons, C. Sun, D. Lohse, Microbubbles and Microparticles are Not Faithful Tracers of Turbulent Acceleration, Phys. Rev. Letters 117 (2016). [46] M. Maxey, J. Riley, Equation of motion for a small rigid sphere in a nonuniform ow, Phys. Fluids 26 (1983) 883{889. [47] R. Gatignol, The Faxen formulas for a rigid particle in an unsteady non-uniform Stokes- ow, J. Mecan. Theor. Appl. 2 (1983) 143{160. [48] H. Gan, J. Chang, J. J. Feng, H. H. HU, Direct numerical simulation of the sedimentation of solid particles with thermal convection, Journal of Fluid Mechanics 481 (2003) 385{411. [49] T. Auton, The lift force on a spherical body in a rotational ow, J. Fluid Mech. 183 (1987) 199{218. [50] S. Balachandar, J. K. Eaton, Turbulent dispersed multiphase ow, Annual Review of Fluid Mechanics 42 (2010) 111{133. [51] B. I. Shraiman, E. D. Siggia, Heat transport in high-rayleigh-number convection, Phys. Rev. A 42 (1990) 3650{3653. [52] E. Calzavarini, Eulerian{lagrangian uid dynamics platform: The ch4-project, Software Impacts 1 (2019) 100002. [53] E. Calzavarini, L. Jiang, C. Sun, Anisotropic particles in two-dimensional convective turbulence, Physics of Fluids 32 (2020) 023305. [54] B. B. Karki, L. P. Stixrude, Viscosity of MgSiO Liquid at Earth's Mantle Conditions: Implications for an Early Magma Ocean, Science 328 (2010) 740. [55] U. R. Christensen, Dynamo scaling laws and applications to the planets, Space Sci. Rev. 152 (2010) 565{590. [56] H. Ni, H. Hui, G. Steinle-Neumann, Transport properties of silicate melts, Rev. Geophys. 53 (2015) 715{744. [57] J. Eaton, J. Fessler, Preferential concentration of particles by turbulence, Int. J. Multiphase Flow 20 (1994) 169{209. [58] R. Mei, Eect of turbulence on the particle settling velocity in the nonlinear drag range, Int. J. Multi. Flow 20 (1994) 273{284. [59] T. Bosse, L. Kleiser, E. Meiburg, Small particles in homogeneous turbulence: Settling velocity enhancement by two-way coupling, Phys. Fluids 18 (2006). [60] S. Clark, F. Spera, D. Yuen, Steady state double-diusive convection in magma chambers heated from below., J. Fluid Mech. (1987) 289{305. [61] S. Grossmann, D. Lohse, Scaling in thermal convection: a unifying theory, Journal of Fluid Mechanics 407 (2000) 27{56. [62] B. D. Marsh, Magma chambers, Annual Review of Earth and Planetary Sciences 17 (1989) 439{472. [63] V. Solomatov, P. Olson, D. Stevenson, Entrainment from a bed of particles by thermal convection, Earth Planet. Sci. Lett. 120 (1993) 387{393. [64] V. Solomatov, D. Stevenson, Nonfractional crystalization of a terrestrial magma ocean, J. Geophys. Res. 98 (1993) 5391{5406. [65] N. M. Qureshi, M. Bourgoin, C. Baudet, A. Cartellier, Y. Gagne, Turbulent transport of material particles: An experimental study of nite size eects, Phys. Rev. Lett. 99 (2007) 184502. [66] D. Brito, J. Aurnou, P. Cardin, Turbulent viscosity measurements relevant to planetary core-mantle dynamics, Phys. Earth Planet. Inter. 141 (2004) 3{8. [67] P. Olson, An experimental approach to thermal-convection in a 2-layered mantle, J. Geophys. Res. 89 (1984) 1293{1301. [68] X. Zhu, V. Mathai, R. J. A. M. Stevens, R. Verzicco, D. Lohse, Transition to the Ultimate Regime in Two- Dimensional Rayleigh-Benard Convection, Phys. Rev. Letters 120 (2018). [69] A. Pandey, J. D. Scheel, J. Schumacher, Turbulent superstructures in rayleigh-b enard convection, Nature Communications 9 (2018) 2118. [70] R. P. J. Kunnen, H. J. H. Clercx, B. J. Geurts, Breakdown of large-scale circulation in turbulent rotating convection, EPL 84 (2008). [71] R. J. A. M. Stevens, H. J. H. Clercx, D. Lohse, Breakdown of the large-scale circulation in Gamma=1/2 rotating Rayleigh-Benard ow, Phys. Rev. E 86 (2012).

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Physics arXiv (Cornell University)

http://www.deepdyve.com/lp/arxiv-cornell-university/settling-of-inertial-particles-in-turbulent-rayleigh-benard-convection-H1mSeSSBMG

Settling of inertial particles in turbulent Rayleigh-Benard convection

Loading next page...

ISSN: 2469-990X
eISSN: ARCH-3341
DOI: 10.1103/PhysRevFluids.5.114304
Publisher site: See Article on Publisher Site

Abstract

Journal

Physics – arXiv (Cornell University)

Published: May 11, 2020

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Settling of inertial particles in turbulent Rayleigh-Benard convection

Settling of inertial particles in turbulent Rayleigh-Benard convection

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Settling of inertial particles in turbulent Rayleigh-Benard convection

Settling of inertial particles in turbulent Rayleigh-Benard convection

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies