Flocking of Active Particles in a Turbulent Flow
Gupta, Anupam;Roy, Amal;Saha, Arnab;Ray, Samriddhi Sankar
2018-12-26 00:00:00
1, 2,y 3,z 4,x Anupam Gupta, Amal Roy, Arnab Saha, and Samriddhi Sankar Ray John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138 Department of Computer Science and Automation, Indian Institute of Science, Bangalore 560012, India Department of Physics, Savitribai Phule Pune University, Pune 411007, India International Centre for Theoretical Sciences, Tata Institute of Fundamental Research, Bangalore 560089, India (Dated: December 27, 2018) We investigate the eect of cooperative interactions in an ensemble of microorganisms, modelled as self-propelled disk-like and rod-like particles, in a three-dimensional turbulent
ow to show
ock- ing as an emergent phenomenon. Building on the work by Choudhary, et al. [Europhys. Lett. 112, 24005 (2015)] for two-dimensional systems, and combining ideas from active matter and tur- bulent transport, we show that non-trivial correlations between the
ow and individual dynamics are essential for the microorganisms to
ock in, for example, a turbulent three-dimensional, marine environment. Our results may have implications especially in the modelling of arti cial microswim- mers in a hostile environment. PACS numbers: 47.63.-b 92.20.Jt 47.27.-i The understanding of collective, cooperative motion of in a turbulent environment by combining ideas from tur- organisms is one of the most important problems in ar- bulent transport [20{22] and active matter [23{26] and eas spanning physical biology, soft matter, and statistical elucidate the critical role of the non-trivial correlations physics. This is largely due to its ubiquity in the natural between the
ow and individual dynamics. Our work world spanning a range of scales from macroscopic (birds shows that the principles of
uid mechanics leave little and sh) to microscopic (bacteria and plankton) organ- choice for organisms but to
ock based on their sizes and isms [1]. Therefore it is reasonable to expect that insights shapes. Thus the inevitable ubiquity of this phenomenon. to this phenomenon is critically related to understanding For several reasons, the collective behaviour of self- self-organised behaviour as well as ecological and evolu- propelled (active) individuals depends on their ambient tionary strategies. Such insights, drawn from a variety medium. Firstly, self propulsion of active particles| of models [2{6], show that combining long-range attrac- which is key to their self-organisation|arises from the tion, short-range repulsion and alignment rules between fact that each individual is able to convert the available self-propelling individuals of model active systems lead free energy into directed motion [25, 26]. Examples in- to self-organised collective dynamics of the individuals. clude eukaryotic cells which utilize the energy produced It thus allows us qualitative as well as quantitative un- by the hydrolysis of available ATP to self-organise and derstanding of the dynamics of real active systems such form tissue during morphogenesis [27] or Janus colloids as cells [7, 8], autophoretic colloids [9], macroscopic, yet with their asymmetric surface chemistry converting the small, organisms [10{12] or droplets [13], synchronised dierence in the chemical potential across their surfaces oscillators [14] and bacterial lattices [15]. and the surrounding medium into self-propulsion [28, 29]. Developing an understanding of
ocking has been Secondly, a
uid medium usually allows for long-ranged largely con ned to the problem of directed motion where hydrodynamic interaction among the active particles, the eects of the ambient medium, air or water|typically which, in turn, in
uences their self-organisation [30{32]. noisy, random and spatio-temporally complex|and the But these interactions, which closely represents the hy- shapes and sizes of the individuals, are ignored. Such an drodynamics of various soft systems such as colloidal sus- assumption for the natural world, where the environment pensions [33, 34], are at Reynolds numbers which are essentially zero. Hydrodynamics at low Reynolds num- can be non-trivial and even turbulent (such as for ma- ber [35] is, however, not applicable to the collective dy- rine life), is an over-simpli cation. Indeed, the few stud- namics of the self-propelled microorganisms in a turbu- ies which have tried to answer this question use either model
ows [16, 17] with mass-less, point-sized organ- lent environment. isms or the relatively more realistic and instructive, but For organisms which move and
ock in air or water, a still academic, case of two-dimensional turbulent
ows na ve guess would be that the
ocks ought to break up with nite-sized spherical particles [18] and, more re- in the presence of such strong perturbations [16]. How- cently,
ocking colloids in a randomly and arti cially ever, observations show that a wide class of organisms disordered environment [19]. These studies, however, are able to overcome such strong perturbations and show strongly suggest the need to study whether collective mo- evidence of collective behaviour [36, 37]. Therefore it is tion can exist in the most general, even if simpli ed, set- important to investigate if the active nature of microor- ting. Therefore in this paper, we now show how and ganisms and their strategy to self-organise can overcome why
ocking emerges|surprisingly|for microorganisms their Stokesian drag and turbulent mixing, to show or- arXiv:1812.10288v1 [cond-mat.soft] 26 Dec 2018 2 for non-spherical particles [20, 21, 40] dr dv A KA [v u(r ; t)] i i i i = v ; = ; (2) dt dt 6a where i is the particle index. Not surprisingly the in- stantaneous orientation is determined by the local geom- etry of the advecting
ow u, namely the symmetric S (strain rate) and antisymmetric (vorticity) tensors of the
uid-velocity-gradients at the particle positions and inertial eects in the rotational dynamics are negligible as shown in previous studies [41, 42]. Inertial eects are however important in the translational degrees of free- dom (Eq. (2)) where the eective drag on the individual depends on the instantaneous orientation in the
ow as well as its shape , the ratio of the semi-major and mi- nor axes of the particles [43, 44]: = 1 is a spherical; < 1 is disk-like (oblate) and > 1 is rod-like (prolate). Thus, the resistance tensor K and the orthogonal trans- FIG. 1. A representative snapshot of a randomly chosen sub- formation matrix A are necessary to take into account set, for clarity, of a population of microorganisms which shows the shape of the microorganism [43, 45] and this leads collective,
ocking behaviour. The blue arrows are the indi- vidual velocity vectors; the thick red arrow is a guide to the to a time-dependent drag coecient, in contrast with eye showing the average direction in which the
ock is moving. the constant drag associated with a spherical particle. (See also the online videos from our simulations [46].) However, for convenience we de ne the average Stokes number St = = [41, 45] to quantify our results as well as compare microorganisms of dierent shapes; is the characteristic small-scale Kolmogorov time-scale ganised,
ocking behaviour in a three-dimensional (tur- of the ambient
uid. Our particles are of course active bulent)
uid environment? And, if so, are there preferred with a self-propelled velocity along the orientation vec- sizes and shapes which lead to
ocking as a spontaneously tor v p ^ . Coupling this with translational equations of 0 i emergent phenomenon in a collection of self-propelled motion, the instantaneous velocity vector of the i-th indi- particles in a turbulent
ow? vidual is a superposition of these two competing eects, We therefore perform detailed numerical simulations of namely v = v + v p ^ : i i 0 i the motion of a suspension of self-propelled active par- Let us nally introduce interactions in a collection of ticles, of dierent shapes and sizes, in three-dimensional such active microorganisms. We use Vicsek-like inter- turbulence. We use a standard pseudo-spectral method actions [23] to ensure that at discrete time intervals T to obtain turbulent solutions of the driven, incompress- each individual i orients itself along the average direction avg ible Navier-Stokes equation and choose parameters to en- n ^ of all individuals N (with the same shape and size) int sure a Taylor-scale Reynolds number Re = 130 which within a radius of interaction r of it. This radius of in- int is consistent with typical marine settings [38]. These teraction is a way to ensure, as is the case in actual living microorganisms, modelled as spheroids, have both rota- organisms, that an individual is not able to see others in tional and translational degrees of freedom and their dy- the population who are physically far away. Of course namics is governed by (a) the drag due to the carrier
ow this assessment of the average orientation cannot be per- which depends on their individual sizes and shapes, (b) fect. To account for this imperfection we introduce an their self-propelled velocity, and (c) their short-ranged avg additive small random noise in the calculation of n ^ alignment interactions with other individuals within a for each individual. This implies neighbourhood. We ensure that the particle sizes are avg smaller than the characteristic Kolmogorov length scales v (t ) = jv (t )jn ^ ; (3) i ? i ? of the
uid. The dynamics of such particles are charac- where t = mT (m is an integer). A similar strategy terised by their (unit) orientation vector p ^ following [39] is used for the re-orientation at discrete time intervals for the vector p ^ . Between successive reorientations, par- dp ^ 1 ticles evolve through the linear, Stokes drag model and :p ^ + [S:p ^ (p ^ :S:p ^ )p ^ ] (1) i i i i i dt + 1 its self-propelled velocity as described. We report results from 3 dierent values of r , namely, 0:03, 0:2, and int and their translational velocity vector, determined from 0:4. (This strategy is also found to be useful for many the linear Stokes drag model (valid for small particles) other purpose such as collective foraging [36].) We also 4 (a) (c) (b) int FIG. 2. Representative log-log pseudo-color plots of the global order parameter as a function of the size and shape for (a) r = 0:03 and v = 0, (b) r = 0:2 and v = 0:2u , and (c) r = 0:4 and v = 1:7u . int 0 rmint 0 int 0 FIG. 2. Global order parameter. Representative log-log pseudo-color plots of the global order parameter as a function of the size and shape for di↵ erent radii of interactions (0.03⇡ ,0.2⇡ ,and 0.4⇡ , from left to right) as well as di↵ erent levels of use three dierent values of the self-propelled velocity: in dierent directions for dierent individuals leading to activity (v =0, the passive case, v =0.2u ,and v =1.7u , from bottom to top). 0 0 ⌘ 0 ⌘ v = 0 (the passive case), v = 0:2u , and v = 1:7u , = 0:0. Hence, since by de nition 0:0 1:0, such 0 0 0 v v where u is the characteristic small-scale (Kolmogorov) a global order parameter quanti es the degree of
ocking uid velocity. Furthermore we use several values of in a species unambiguously. is clear there is no ordering when particles are essentially preferentially concentrate for a range of Stokes numbers. and St (which corresponds to particles, or microorgan- passive but still allowed to interact over the same length A convenient quantification of this inhomogeneous dis- Figure 2 shows representative pseudo-color plots of isms with sizes which vary between 1m to 50m) and scale. This is of course consistent with the picture that tribution of particles in a flow is through the correla- as a function of and St (on a log scale) for dierent r int perform simulations with N = 50; 000 particles for every emerged when we looked at the global order parameter tion dimension D which is measured via the probabil- and v (see gure caption). 2 These results are intriguing ; St, r , and v . int int 0 . (A similar definition is possible for the orientation ity of finding two particles within a distance r, namely, v for several reasons. For individuals which are not active, < D It is useful to begin with an intuitive picture of whether 2 vectorˆp; we have checked that the fluctuations are far P (r) ⇠ r . For tracers, which are space filling D =3 i.e., v = 0, microorganisms
ock only when they are 2 there is any evidence of collective motion in our model. stronger for the orientation vector as discussed at the end (in 3 dimensional flows). However for non-tracers (iner- suciently large, rod-like and with a very large radius FIG. 2. Global order parameter. Representative log-log pseudo-color plots of the global order parameter as a function This is best seen online in videos [46] of the time evolu- of this paper.) tial particles, St 6= 0), as the finiteness of the particles of interaction. This clearly shows that in the absence of of the size and shape for di↵ erent radii of interactions (0.03⇡ ,0.2⇡ ,and 0.4⇡ , from left to right) as well as di↵ erent levels of tion of a collection of such microorganisms with dierent become important, D has a non-monotonic dependence activity, it is extremely hard for microorganisms which activity (v =0, the passive case, v =0.2u ,and v =1.7u , from bottom to top). 0 0 ⌘ 0 ⌘ shapes, Besfor izes, e w activit e addr yes (sv t)hand e quranges estion of of in thteraction e relevanc (re of)our 0 int are on smal thl,e slender Stokes nor um disk-shap bers: It d ed, ecrto easoevsercome from Dturbulen =3 witth in- with finrandom dings to initial naturep , osition it is tim s.e These to prov sim ideulations a theoresho tical w ex- mixing creas and ingdrag Stokto es self-organise. numbers and In reac this hesparameter a minimaspace, for Stokes conp vincingly lanation hof owth
o iscking pheneom merges enon.only In when recentconditions years, there evennu ifman berin s c dividual lose to 1aligns beforewith incre its asineigh ng agai bour, n toev sat enutu- rate at is clear there is no ordering when particles are essentially preferentially concentrate for a range of Stokes numbers. are right in terms of not only shapes and sizes but also have been several studies on the dynamics of small, pas- ally that direction gets randomised due to the turbulent D = 3 for Stokes numbers much larger than 1 [34]. It passive but still allowed to interact over the same length A convenient quantification of this inhomogeneous dis- v and r . Indeed, for the cases where they do
ock, a sive, spherical particles in turbulent flows. These stud- 0 int medium and therefore it cannot continue with the di- is this non-monotonic (as a function of the Stokes num- scale. This is of course consistent with the picture that tribution of particles in a flow is through the correla- snapshot of the velocity vectors of the individuals (Fig. 1) ies have shown that finite-sized inertial particles tend to rection over time which is necessary for the
ocking to emerged when we looked at the global order parameter tion dimension D which is measured via the probabil- clearly re
ect the level of ordering in our model. For clar- emerge. . (A similar definition is possible for the orientation ity of finding two particles within a distance r, namely, int ity, in Fig. 1 we only show a randomly chosen subset of < D vectorˆp; we have checked that the fluctuations are far P (r) ⇠ r . For tracers, which are space filling D =3 However as soon as we turn on activity, i.e., v 26= 0 the total N rod-like ( = 2:0) and inertial (St = 2:0) stronger for the orientation vector as discussed at the end (in 3 dimensional flows). However for non-tracers (iner- but still small (Fig. 2b), we obtain non-zero values of microorganisms in the
ow. of this paper.) tial particles, St 6= 0), as the finiteness of the particles FIG. 2. Global order parameter. Representative log-log pseudo the -colorder or plotparameter s of the glofor bal all ordsp er ecies param only eter for as alarge functsiizes on With this picture in mind, and visual evidence of
ock- of the size and shape for di↵ erent radii of interactions (0.03⇡ ,0.2⇡ ,a bn ecdom 0.4e⇡ i,mfro por mtan leftt,tD o righhas t)aasnon we-lm l aon s doti↵ on ere icntdelepveenlsdeonfce (St > 1). For smaller-sized 2 microorganisms for a given ing (depending on the type of microorganism), we cannot Before we address the question of the relevance of our activity (v =0, the passive case, v =0.2u ,and v =1.7u , from bottom to top). 0 0 ⌘ 0 ⌘ on the Stokes numbers: It decreases from D =3 with in- radius of interaction, the shape plays a crucial role: Rods refrain anymore from making our study quantitative. A findings to nature, it is time to provide a theoretical ex- creasing Stokes numbers and reaches a minima for Stokes are more likely to show collective behaviour than disks convenient way to quantify the degree of
ocking, is to planation of this phenomenon. In recent years, there numbers close to 1 before increasing again to saturate at because, as has been known, correlation time-scales for measure the global order parameter [23] is clehar avtehb ereeenissneoveor raldesrtiundgiewshon entphar e tdiyclneam s ariceseof ssesnmtial all,lypas- preferentially concentrate for a range of Stokes numbers. D = 3 for Stokes numbers much larger than 1 [34]. It rods are considerably longer than disks [45]. Understand- * + sive, spherical particles in turN bulent flows. These stud- passive but still allowed to interact pover the same length A convenient quantification of this inhomogeneous dis- is this non-monotonic (as a function of the Stokes num- X ably, as the r and v increase (e.g., Fig. 2c), the degree int 0 ies have shown that finite-sized inertial particles tend to scale. This is of course consistent with the picture that tribution of particles in a flow is through the correla- = v ^ (4) v j of
ocking enhances for all species. Nevertheless, it is emerged when we looked at the global order parameter tion dimension D which is measured via the probabil- j=1 clear from our results that both size and shape matter| . (A similar definition is possible for the orientation ity of finding two particles within a distance r, namely, in a non-monotonic and complicated way|decisively in where the angular brackets denote a time-average (after < D vectorˆp; we have checked that the fluctuations are far P (r) ⇠ r . For tracers, which are space filling D =3 determining the ability of a species to overcome turbu- the initial transients in their motion have died down) and stronger for the orientation vector as discussed at the end (in 3 dimensional flows). However for non-tracers (iner- lent mixing and drag to swarm. Indeed for a signi cant v ^ is the unit velocity vector of the j-th individual. This of this paper.) tial particles, St 6= 0), as the finiteness of the particles range of values of , St, v , and r , we do nd evidence 0 int order parameter ought to vary from species to species of bec ofom near e impperfect ortant
o , D cking has (a n! on1-m :0). onotonic dependence dierent shapes and sizes as well as depend on the level Before we address the question of the relevance of our on the Stokes numbers: It decreases from D =3 with in- of activity and radius of interaction. Should organisms This is the rst clear evidence of how collective be- findings to nature, it is time to provide a theoretical ex- creasing Stokes numbers and reaches a minima for Stokes ock, then all velocity vectors must point in the same haviour can emerge in a collection of microorganism in a planation of this phenomenon. In recent years, there numbers close to 1 before increasing again to saturate at direction yielding = 1:0. On the other hand, if there complex, random turbulent medium. Before we provide have been several studies on the dynamics of small, pas- D = 3 for Stokes numbers much larger than 1 [34]. It is no ordering, then the velocity vectors ought to point a theoretical explanation of this emergent phenomenon, sive, spherical particles in turbulent flows. These stud- is this non-monotonic (as a function of the Stokes num- ies have shown that finite-sized inertial particles tend to 4 ing, D = 3 (in 3 dimensional
ows). However for non- tracers (St 6= 0), it decreases from D = 3 with increasing St and reaches a minima for St = O(1), before increasing again to saturate at D = 3 for St 1 [21]. It is this non- monotonic (as a function of the Stokes number) nature of preferential concentration of heavy, inertial particles which we show is central to
ocking of our active parti- cles. Hence, the particle dynamics ensure that for certain values of and St more organisms are forced, mechani- cally, to be close to each other and hence an enhancement in the degree of
ocking. Therefore if the microorganisms are active, such preferential concentration plays a dom- inant role in the emergence of
ocking. In the absence of preferential concentration, the chances of a large frac- tion of microorganisms clustered close enough for their mutual alignment-interactions to become eective would be minimal. FIG. 3. Representative plot of the nematic order, as a func- Our ndings could be of importance for fabrication tion of size St, for dierently shaped microorganisms for and design of arti cial micro-swimmers. But are they r = 0:2. We show results for both active (v = 0:2u ) as int 0 equally relevant for observations in nature? Typical well as (inset) passive (v = 0) microorganisms. marine environments are known to be turbulent with a Reynolds number comparable to the ones we use in our simulations and a Kolmogorov scale of the or- it is useful to characterise one additional aspect, namely der of a few mm [36, 38]. In such an environment, the nematic order [47], of these
ocks. This is done most microorganisms|not necessarily spherical|such as zoo- conveniently through the tensorial order parameter plankton often show collective behaviour [48]. Recent measurements suggest that this class of organisms have a Q = v ^ v ^ 1 ; (5) i j variety of sizes which could range from 1mm to 10mm and, hence, less than [36, 48]. This scale separation ensures that our modeling of microorganisms through a whence we can calculate the scalar order parameter, asso- v v combination of the linear Stokes drag model [20{22], for ciated with the nematic order, S (x) = n n Q , where the translational degrees of freedom and the Jeery equa- n (x) are the local symmetry directions of the elds. tion [39, 45, 49, 50], for the rotational degrees of freedom, Since we are interested in the global behaviour of this sys- is valid. Additionally, the associated response time scale, tem, we calculate the global nematic order = dxS . v v the Stokes time, 1s 1s, of such microorganisms In Fig. 3 we show a representative plot of , for a given v p when compared to the Kolmogorov time scale of the radius of interaction and self-propelled velocity as well ocean environment leads to Stokes numbers of the same as for, in the inset, the same r but for v = 0, as a int 0 order as we have used in our study [48]. function of the Stokes number and for dierent shapes . Although we have not obtained perfect nematic or- These arguments suggest that our theoretical frame- dering, we do see that a certain level of order emerges work, even with its limitations (see below), is a rele- only for Stokes numbers larger than 1. From the inset, it vant model for small organisms in a marine environ- is clear there is no ordering when particles are essentially ment. Our results show that collective, ordered mo- passive but still allowed to interact over the same length tion in active systems is an emergent phenomenon which scale. This is of course consistent with the picture that can spontaneously occur even in a hostile environment emerged when we looked at the global order parameter where the range of interactions (here, quanti ed by r ) int . (A similar de nition is possible for the orientation for a collection of individuals is restricted. These re- vector p ^; we have checked that the
uctuations are far sults show that purely mechanical principles related to stronger for the orientation vector as discussed at the end the dynamics of nite-sized particles in turbulence are of this paper.) critical in forcing certain organisms to
ock: Given the We know that nite-sized, passive, inertial particles right size and shape, microorganisms are brought in much tend to preferentially concentrate for a range of Stokes closer proximity to each other allowing them to behave numbers. A convenient quanti cation of this inhomoge- cooperatively and
ock. Indeed recent studies of pas- neous distribution of particles in a
ow is through the cor- sive, non-interacting, non-spherical particles shed light relation dimension D which is measured via the proba- on the correlation between particle trajectories and
ow bility of nding two particles within a distance r, namely, directions [45, 49, 50]; as particles become interacting < D P (r) r . For tracers (St = 0), which are space ll- and self-propelled the competing eects between the pas- 5 sive and active cases lead to our very interesting|and 1999. [2] C. M. Topaz, A. L. Bertozzi and M. A. Lewis, Bull. Math. surprising|discovery that shape and size both matters Biol. 68 1601 (2006). for stabilising model
ocks in turbulent
ows. Without [3] R. Eftimie, G. de Vries G, M. A. Lewis and F. Lutscher, these underlying principles of
uid mechanics a random, Bull. Math. Biol. 69, 1537 (2007). chaotic environment would have pushed individuals far [4] I. Giardina, HFSP J. 2, 205 (2008). apart leading to a break up of
ocks. Hence, by bringing [5] F. Alaimo, S. Praetorius and A. Voigt, New J. Phys. 18, together basic principles of turbulent transport and ac- 083008 (2016). tive matter, we have, for the rst time, shown how model [6] J. Dunkel, S. Heidenreich, M. Baer and R. E. Goldstein, New J. Phys. 15, 045016 (2013). ocks can form in a collection of self-propelled individuals [7] E. Mehes and T. Vicsek, Integr. Biol. 6, 831 (2014). moving in a turbulent
ow. Such cooperative behaviour [8] X. Chen, X. Dong, A. Be'er, H.L. Swinney and H.P. has recently been reported for dry, granular systems [51] Zhang, Phys. Rev. Lett. 108, 148101 (2012). but not for the complex system that we report here. More [9] I. Buttinoni, J. Bialk e, F. Kummel, H. Loewen, C. pertinently, Durham, et al. [37], considered modelled ac- Bechinger, Phys. Rev. Lett., 110, 238301 (2013). tive tracers to understand the observed patchiness of mi- [10] J. Buhl, D.J.T. Sumpter, I. D. Couzin, J.J. Hale, E. De- croorganisms such as phytoplankton: However, the idea spland, E.R. Miller and S.J. Simpson, Science 312, 1402 (2006). of collective motion and the role played by inertia as well [11] I. D. Couzin, J. Kranse, R. James, G. D. Runton and N. as the translational and rotational degrees of freedom was R. Franks, J. Theor. Biol. 218, 1 (2002). largely ignored in this and other studies. [12] D. H. Kelley and N.P. Ouellette, Sci. Rep. 3, 1073 (2013). Before we conclude, it is essential to understand some [13] S.Thutupalli, D. Geyer, R. Singh, R. Adhikari, and of the limitations of our model. Our model cannot cap- Howard A. Stone, Proc. Natl. Acad. Sci. U.S.A, 115 (21), ture the more dramatic instances of
ocking in nature 5403-5408 (2018). which involve macroscopically large organisms such as [14] K. P. O'Keee, H. Hong and S. H. Strogatz, Nat. Comm. 8, 1504 (2017). birds. This class of phenomena should be studied with [15] H. Wioland, F. G. Woodhouse, J. Dunkel and R. E. Gold- a similar
uid mechanics approach in future by resolving stein, Nat. Phys. 12, 341 (2016). boundaries and developing ideas for large structure-
uid [16] N. Khurana and N. Ouellette, New J. Phys. 15, 095015 interactions. Furthermore, the role of hydrodynamic in- (2013). teractions in
ock stability ought to be investigated sys- [17] A. W. Baggaley, Phys. Rev. E 91, 053019 (2015). tematically within the present model. Finally, a more [18] A. Choudhary, D. Venkataraman, and S. S. Ray, Euro- systematic study of nite (population) size eects and phys. Lett. 112, 24005 (2015) [19] A. Morin, N. Desreumaux, J.-B. Caussin and D. Bartolo, the eect of intrinsic noise as well as the level of activity Nat. Phys. 13, 63 (2017). is left for future work. [20] J. Bec, Phys. Fluids 15, 81 (2003); J. Fluid Mech. 528, We hope that this novel|yet simple|mechanical ap- 255 (2005). proach, bringing together already well-established con- [21] J. Bec, L. Biferale, M. Cencini, A. Lanotte, S. Musacchio, cepts in dierent areas, is an important ingredient in and F. Toschi, Phys. Rev. Lett. 98, 084502 (2007). explaining why
ocking is ubiquitous, and is as much [22] E. W. Saw, G. P Bewley, E. Bodenschatz, S. S. Ray, and a strategy as it is forced. Such an approach could also J Bec Phys. Fluids 26 (11), 111702 (2014). [23] T. Vicsek, A. Czir ok, E. Ben-Jacob, I Cohen, and O. lead to new ideas beyond the present study such as de- Shochet, Phys. Rev. Lett. 75 1226 (1995). veloping more realistic predator-prey models in complex [24] T. Vicsek and A. Zafeiris, Physics Reports, 517, 3-4, pp. environment. 71(2012). SSR acknowledges the support of the DAE, Indo{ [25] M. C. Marchetti, J.-F. Joanny, S. Ramaswamy, T. B. French Center for Applied Mathematics (IFCAM) and Liverpool, J. Prost, M. Rao, and R. A. Simha, Rev. Mod. the Airbus Group Corporate Foundation Chair in Math- Phys. 85 (3), 1143(47), (2013). ematics of Complex Systems established in ICTS. AR [26] C. Bechinger, R. Di Leonardo, H. L owen, C. Reichhardt, G. Volpe and G. Volpe, Rev. Mod. Phys. 88 (4), 045006 and SSR acknowledges the support of the DST (India) (50), (2016). project ECR/2015/000361. The simulations were per- [27] D. W. Thompson, On growth and form, Cambridge Univ. formed on the cluster Mowgli and workstations Goopy Press (1992). and Bagha at the ICTS-TIFR. AS acknowledges start-up [28] J. Palacci, S. Sacanna, A. P. Steinberg, D. J. Pine, P. M. grant (No.4-5 (206-FRP)/2015(BSR)) from UGC. Chaikin, Science 339, 936 (2013). [29] W. F. Paxton, K. C. Kistler, C. C. Olmeda, A. Sen, S. K. St. Angelo, Y. Cao, T. E. Mallouk, P. E. Lammert, and V. H. Crespi, JACS, 126, 41, pp. 13424-13431(2004). [30] R. Singh and R. Adhikari Phys. Rev. Lett. 117, 228002 anupam1509@gmail.com (2016). [31] A. M. Menzel, A. Saha, C. Hoell, and H. L owen, J. Chem. amalroy@iisc.ac.in sahaarn@gmail.com Phys. 144, 024115 (2016). [32] M. Hennes, K. Wol, and H. Stark, Phys. Rev. Lett. 112, samriddhisankarray@gmail.com [1] J. K. Parrish and L. Edelstein-Keshet, Science 284 99 238104 (2014). 6 [33] P. Mazur, Physica 110A, 128 (1982). 1. https://youtu.be/yPNb6SKRNVI [Disk-like Microor- [34] P. Mazur and W. Van Saarloos, Physica, 115, 21 (1982). ganisms showing Collective Behaviour] [35] E. M. Purcell, Am. J. Phys. 45, 3 (1977). 2. https://youtu.be/VR9NqAC-kzk [Disk-like Microor- [36] R. Stocker, Science 338, 628 (2012); H. Marcos, H. C. Fu, ganisms not showing Collective Behaviour] T. R. Powers, R. Stocker, Proc. Natl. Acad. Sci. U.S.A. 3. https://youtu.be/s7lh-s2NTZk [Rod-like Microor- 109, 4780 (2012). ganisms showing Collective Behaviour] [37] W. M. Durham, E. Climent, M. Barry, F. De Lillo, G. 4. https://youtu.be/6z0ky__ZZ9s [Rod-like Microor- Boetta, M. Cencini, and R. Stocker, Nat. Comm. 4, ganisms not showing Collective Behaviour]. 2148 (2013). [47] F. Ginelli, F. Peruani, M B ar, and H. Chat e, Phys. Rev. [38] J. Jim enez, Sci. Mar., 61 47 (1997). Lett. 104, 184502 (2010). [39] G. B. Jeery, Proc. R. Soc. A 102, 161 (1922). [48] J. Yen and E. A. Bundock, Aggregate behavior in zoo- [40] M. R. Maxey and J. J. Riley, Physics of Fluids 26, 883 plankton: Phototactic swarming in four developmen- (1983). tal stages of Coullana canadensis (Copepoda, Harpacti- [41] L. Zhao, N. R. Challabotla, H. I. Andersson, and E. A. coida), in Animal Groups in Three Dimensions, ed. J. Variano, Phys. Rev. Lett. 115, 244501 (2015). K. Parrish and W. M. Hamner (Cambridge University [42] C. Marchioli, L. Zhao, and H. I. Andersson, Phys. Fluids Press) pp. 143-162 (1997); J. W. Ambler, Hydrobiologia, 28, 013301 (2016). 480 155164 (2002). [43] P. H. Mortensen, H. I. Andersson, J. J. J. Gillissen, and [49] A. Pumir and M. Wilkinson, New J. Phys. 13 093030 B. J. Boersma, Phys. Fluids 20, 093302 (2008). (2011); A. Gupta, D. Vincenzi and R. Pandit, Phys. Rev. [44] N. R. Challabotla, L. Zhao, and H. I. Andersson, J. Fluid E 89 021001(R) (2014). Mech. 766, R2 (2015). [50] S. Parsa, E. Calzavarini, F. Toschi, and G. A. Voth Phys. [45] A. Roy, A. Gupta, and S. S. Ray, Phys. Rev. E, 98, Rev. Lett. 109, 134501 (2012). 021101(R) (2018). [51] N. Kumar, H. Soni, S. Ramaswamy, and A. K. Sood, Nat. [46] Online videos from our simulations: Comm. 5, 4688 (2014).
http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.pngCondensed MatterarXiv (Cornell University)http://www.deepdyve.com/lp/arxiv-cornell-university/flocking-of-active-particles-in-a-turbulent-flow-TjBg0K10MX