Pith. sign in

REVIEW 2 major objections 5 minor 44 references

Collective ballistic motion explains fast aggregation in adhesive active matter

T0 review · 2 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Strong self-alignment in adhesive active matter triggers collective ballistic aggregation, in which the average cluster mass grows as M(t) ~ t², faster than diffusion- or persistence-limited aggregation.

desk verdict Nice simulation study with a plausible persistence-length mechanism for z≈2 in adhesive active matter, but the analytic derivation of the headline exponent has a real gap at γ=0. read the letter →

arxiv 2508.11793 v2 pith:XCEGH6IY submitted 2025-08-15 cond-mat.soft physics.bio-phphysics.comp-ph

Emanuel F. Teixeira , Pablo de Castro , Carine P. Beatrici , Heitor C. M. Fernandes , Leonardo G. Brunnet This is my paper · ORCID
classification cond-mat.softphysics.bio-phphysics.comp-ph
keywords adhesiveactivematteraggregationkineticscollectivemotionflockingtransitionSmoluchowskicoagulationself-alignmentclusterpersistencelengthcell
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is trying to establish why sticky, self-propelled particles—motile cells in tissue formation being the motivating case—can aggregate much faster than classical coagulation theory predicts, with the mean cluster mass growing as t² rather than $t^{{1/2}}$ or t. The answer it proposes is a flocking transition: strong self-alignment makes each cluster move coherently, and the distance a cluster travels while keeping its direction grows linearly with cluster mass, while the typical gap between clusters grows only as the square root of mass. As soon as the persistence length overtakes the intercluster gap, collisions become ballistic and the system enters a collective ballistic aggregation regime that is asymptotically self-sustaining. The same persistence-length criterion, supported by a generalized Smoluchowski coagulation theory, organizes all measured exponents (z from 1/2 to 2) into a regime diagram in the plane of self-alignment strength and single-particle persistence.

What carries the argument

The load-bearing comparison is between two lengths: the cluster persistence length lp = Vcτp, the distance a cluster travels before its direction decorrelates, and the mean intercluster distance lc = (M/ρ)^{1/2} set by the global density. Self-alignment changes the mass scalings of single-cluster dynamics—strong alignment drives the polar order parameter toward 1, saturates Vc at v0 (γ = 0), and makes τp grow linearly with M (ν = 1)—so lp ∝ M while lc ∝ $M^{{1/2}}$, and the criterion 2lp/lc > 1 marks entry into collective ballistic aggregation, with lp/lc ∝ $M^{{1/2}}$ making that regime self-sustaining. The companion machinery is a generalized Smoluchowski coagulation theory whose kernels encode these scalings, giving z = 1/(1−α) for diffusion-limited aggregation and z = 2/(1−2γ) for ballistic aggregation in two dimensions.

What would settle it

Measure the coalescence rate of two equal-mass clusters moving at the same speed in random directions in the strong-alignment regime; if that rate tends to zero as the speed difference vanishes, the t² law lacks the required source of relative motion and the collective-ballistic explanation, as written, fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that the anomalous aggregation exponent z ≈ 2 (mean cluster mass growing as M(t) ~ t²) is produced by a flocking transition inside clusters, not by the persistence of individual particles. Single-cluster simulations in the strong-alignment limit show that the cluster speed becomes independent of mass, Vc = v0 (exponent γ = 0), while the persistence time grows linearly with mass, τp ∝ M (exponent ν = 1). The persistence length lp = Vcτp therefore scales as M, while the average intercluster distance lc = (M/ρ)^{1/2} scales as $M^{{1/2}}$, so the ratio lp/lc grows as $M^{{1/2}}$; once the criterion 2lp/lc > 1 is met, clusters collide ballistically and keep doing so as they grow, making the t² regime asymptotic rather than transient. A generalized Smoluchowski coagulation equation with diffusion kernel ∝ (D(M)+D(M′))(R+R′)^{d−2} and ballistic kernel ∝ |Vc(M)−Vc(M′)|(R+R′)^{d−1} yields the exponent formulas z = 1/(1−α) and z = 2/(1−2γ) in d = 2, reproducing the simulations and uniting diffusion-limited (z ≈ 1/2–1), non-collective ballistic (z ≈ 1), and collective ballistic (z ≈ 2) aggregation in a single picture.

Load-bearing premise

The t² law rests on the unstated premise that clusters of equal mass and equal speed still meet at a nonzero rate because their motion directions differ; if collisions required a difference in speed, equal-speed clusters would never coalesce and the predicted fast regime would not follow.

Editorial extensions

If this is right

  • Cell-sorting and tissue experiments that report aggregation exponents between 1 and 2 can be interpreted as sitting in the crossover between diffusion-limited and collective ballistic aggregation, without invoking new physics.
  • The criterion 2lp/lc > 1 is directly measurable: tracking cluster centroids to obtain Vc(M) and τp(M) predicts the aggregation regime of a given adhesive active system.
  • Because lp/lc keeps growing as M^{1/2} in the collective regime, the t² law is self-stabilizing: larger clusters are even further into the ballistic regime, so no parameter tuning is needed to sustain fast aggregation.
  • Systems in collective ballistic aggregation develop a power-law cluster-mass distribution with dynamical scaling (exponent λ ≈ 1.1), whereas diffusion-limited aggregation retains a characteristic cluster size, giving experiments a statistical fingerprint that distinguishes the regimes.
  • High single-particle persistence without alignment produces a long-lived transient with z ≈ 1 before the crossover to diffusion-limited behavior, so early-time exponents in biological assays need not reflect the asymptotic regime.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves implicit that the ballistic kernel written in the Supplement, |Vc(M) − Vc(M′)|, vanishes when the cluster speed is mass-independent (γ = 0); obtaining z = 2 from the written equations requires the added premise that equal-speed clusters still approach each other with nonzero relative velocity because their directions differ.
  • With that premise made explicit, the Smoluchowski formalism extends naturally to an angle-averaged kernel proportional to Vc, which would give an explicit d-dimensional prediction for z(γ, d) instead of the collinear-kernel formula z = d/(1 − dγ).
  • A testable extension suggested by the mechanism: reducing adhesion so that clusters can fragment should cut off the linear growth of lp with M and lower the measured exponent—the paper lists fragmentation as future work.
  • The persistence-length criterion maps onto an experimental protocol: measuring Vc and τp for cell aggregates of increasing size should show lp(M) turning linear at the same alignment strength at which the aggregation exponent jumps toward 2.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The manuscript studies aggregation of adhesive active Brownian disks with self-alignment, varying the alignment strength J and the Péclet number Pe. It identifies three kinetic regimes: diffusion-limited cluster aggregation (DLCA, z between 1/2 and 1), non-collective ballistic aggregation (z ≈ 1), and a collective ballistic aggregation (CBA) regime with z ≈ 2. Single-cluster simulations are used to extract mass-dependent cluster speed, persistence time, and persistence length, and a criterion comparing persistence length with intercluster distance is proposed to locate the crossover to ballistic aggregation. A generalized Smoluchowski coagulation theory with diffusion- or velocity-based kernels is then used to express the aggregation exponent z in terms of measured single-cluster exponents, and the results are summarized in a Pe–J regime diagram.

Significance. If the CBA mechanism is correct, the paper provides a simple and appealing explanation for anomalously fast aggregation exponents observed in adhesive active matter: flocking makes the cluster persistence length grow linearly with mass while the intercluster distance grows only as the square root of mass, so large clusters move in a collective ballistic regime with M(t) ~ t^2. The paper's strengths are its systematic simulations, the clean single-cluster observable lp(M), the dynamical-scaling collapse of P(M,t), and the unified regime diagram. The main theoretical derivation of the headline exponent z=2, however, has a gap at the precise point γ=0, because the ballistic kernel written in the Supplement vanishes identically for mass-independent cluster speed. The result is likely repairable by adding an explicit orientational-disorder premise, but as written the central claim is not derived from the stated equations.

major comments (2)
  1. [Supplement, Eqs. (S24)–(S31); main text after Eq. (S31)] The ballistic kernel is written as K(M,M′) = B |M^γ − M′^γ| (M^{1/df} + M′^{1/df})^{d−1}. For the CBA regime the paper uses γ=0, corresponding to mass-independent cluster speed. At γ=0 the factor |M^γ − M′^γ| is identically zero, so I2 in Eq. (S27) vanishes, C2=0, and Eq. (S26) gives d⟨M⟩/dt = 0 rather than ⟨M⟩ ~ t^2. Consequently z = 2/(1−2γ) evaluated at γ=0 is not derivable from the kernel as stated. The intended physical picture must be that clusters have equal scalar speeds but randomly oriented velocities, so the mean relative speed is a nonzero, mass-independent constant. With that premise the kernel becomes K ~ (R+R′)^{d−1}, which for d=df=2 is homogeneous of degree 1/2 and indeed gives z=2. The authors should state this orientational-disorder premise explicitly and revise Eq. (S24) accordingly, since the current text presents γ=0 as a sufficient condition.
  2. [Main text, 'To explain z...' paragraph and Fig. 3] The agreement between measured and predicted aggregation exponents is a consistency check between two observables of the same simulation, not an independent prediction. The exponents γ, ν, and α are extracted from single-cluster simulations of the same model and then inserted into the Smoluchowski scaling formulas to obtain z. This is acceptable as a scaling analysis, but the text should be more precise: it says 'Smoluchowski theory predicts z=2', whereas in fact the theory uses the measured mass-dependence of cluster motion as input. Please clarify which ingredients are measured and which are derived, so that readers can judge the explanatory content of the framework.
minor comments (5)
  1. [Main text, after 'We obtain z = 2/(1−2γ)'] The word 'yelding' should be 'yielding'.
  2. [Supplement, Figs. S1 and S2] The captions of Figs. S1 and S2 appear inconsistent with the surrounding text: the text says Fig. S1 is for low Pe and Fig. S2 for high Pe, but the Fig. S1 caption states 'high Péclet number regime (Pe=10^3)'. Please correct the labeling.
  3. [Fig. 3] The measured exponents γ, ν, and α are central inputs to the theory, yet the panels in Fig. 3 show no error bars or fit ranges. Adding uncertainty estimates or at least stating how the power-law fits were performed would strengthen the quantitative claims.
  4. [High persistence, J=0 paragraph] The statement that masses M < 225.73 are in the ballistic regime and larger masses follow DLCA is based on the 2lp/lc criterion, but the M(t) data in Fig. 4a do not show a visible crossover in the displayed time window. A quantitative comparison of the predicted crossover mass with the simulation time scale would make this argument more convincing.
  5. [General] There is no data- or code-availability statement; providing simulation details or a reproducibility statement would be helpful, especially because several claims (robustness to packing fraction, robustness to alignment rule) are mentioned without a figure.

Circularity Check

2 steps flagged · score 6.0 of 10

The z=2 derivation evaluates a ballistic kernel that vanishes identically at γ=0, and the analytic theory inserts measured single-cluster exponents into standard Smoluchowski scaling; the central analytic claim is not derived from the equations as written.

  1. other [Supplemental Material, 'Ballistic aggregation', Eqs. (S24)-(S29); main text paragraph 'For J ≥ 0.8...']
    "K(M, M′) = B |V (M ) − V (M ′)| (R(M ) + R(M ′))d−1 = B |M γ − M ′γ| (M 1/df + M ′1/df )d−1. ... For clusters composed of particles with uncorrelated self-propulsions, γ = −1/2, whereas for fully aligned clusters, γ = 0. ... For J ≥ 0.8, ballistic aggregation occurs with mass-independent cluster speed ( γ = 0), yelding z = 2, consistent with our simulations (Fig. 4a)."

    At γ = 0, the relative-speed factor |M^γ − M′^γ| in Eq. (S24) is identically zero for every pair of masses, so the integral I2 in Eq. (S27) vanishes, giving C2 = 0 and d⟨M⟩/dt = 0 in Eq. (S26), not a growing t^2 law. The paper nevertheless evaluates the asymptotic exponent z = 2/(1 − 2γ) at γ = 0, silently replacing the zero prefactor with a nonzero constant. That constant can only come from unstated orientational disorder (equal-speed clusters moving in different directions), which appears neither in Eq. (S24) nor in the stated premise of mass-independent cluster speed. The headline z = 2 is therefore not a consequence of the written equations; the t^2 growth is imported by hand.

  2. fitted input called prediction [Main text, 'Model' section and Conclusions; Fig. 3 caption]
    "we will assume power-law dependencies in our theory below in order to capture the limiting values of z, as well as its trend, as J is varied. ... For J ≥ 0.8, ballistic aggregation occurs with mass-independent cluster speed ( γ = 0), yelding z = 2, consistent with our simulations (Fig. 4a). ... With mass-independent cluster speed ( γ = 0, Fig. 3a), Smoluchowski theory predicts z = 2."

    The exponent γ = 0 is measured from single-cluster simulations of the same model (Fig. 3a) and then inserted into the standard Smoluchowski scaling formula z = 2/(1 − 2γ) to obtain z = 2, which is compared with aggregation simulations of the same model. This is a consistency check between two simulation analyses of the same particle model, not an independent first-principles prediction of the aggregation exponent. The paper is transparent about assuming the power-law scalings, so the circularity is partial: the Smoluchowski scaling relation is standard and the persistence-length mechanism is separately motivated, but the advertised theory 'predicts' an exponent whose key input (γ = 0) is itself fitted from the model.

full rationale

The numerical study is essentially self-contained: the model is simulated, exponents are measured, and the regime diagram is constructed from those measurements. The self-citations to prior work by the same group (refs. [10] and [23]) are contextual and not load-bearing: the l_c = sqrt(M/ρ) estimate and the observation that alignment modifies the diffusivity-mass relation are supporting background rather than the argument's foundation. No uniqueness theorem or machine-checked external result is invoked to force the paper's choice. The serious problem is internal to the analytic derivation: the ballistic kernel written in Eq. (S24) is exactly zero when γ = 0, the value the main text identifies with the collective ballistic regime, so Eq. (S26) yields d⟨M⟩/dt = 0 unless an unstated nonzero relative-speed term from orientational disorder is added. Evaluating the homogeneous scaling exponent z = 2/(1 − 2γ) at γ = 0 hides this vanishing prefactor. In addition, the analytic theory's inputs (γ, ν, α) are measured from single-cluster simulations of the same model, making the subsequent 'Smoluchowski theory predicts z = 2' a consistency check rather than an independent derivation. The crossover criterion 2l_p/l_c > 1 is a heuristic comparison, not circular by construction, though it is calibrated on the same data. Overall, the empirical aggregation phenomenology is credible and self-contained, but the central analytic claim that collective ballistic aggregation with γ = 0 yields z = 2 is not derivable from the paper's own equations as written; the derivation requires an additional physical premise that is never stated, and the exponents fed into the scaling theory are fitted inputs. This warrants a score of 6 rather than a finding of no circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central derivation is an internal consistency story: single-cluster transport exponents are measured and then fed into standard coagulation scaling formulas to reproduce the aggregation exponents. No new physical entity is introduced. The main ledger costs are the assumed power-laws and the unstated relative-velocity premise in the ballistic kernel.

free parameters (3)
  • cluster speed exponent γ = from -1/2 (low J) to 0 (high J)
    Obtained by power-law fits to single-cluster MSD curves (Fig. 3a); used in z = 2/(1 - 2γ).
  • cluster persistence time exponent ν = from 0 (low J) to ≈1 (high J)
    Obtained from fits of τp versus M (Fig. 3b); feeds δ = γ + ν and the criterion 2lp/lc.
  • cluster diffusion exponent α = from -1 (low J) to 1 (high J)
    Derived as α = 2γ + ν from fitted exponents; also independently measured (Fig. 3c); used in z = 1/(1 - α).
assumptions (5)
  • domain assumption Smoluchowski mean-field coagulation equation with spatially homogeneous cluster concentrations.
    Supplement Eqs. (S4)-(S5) assume spatial uniformity; neglected correlations could alter exponents in low-density 2D aggregation.
  • domain assumption Dynamical scaling ansatz P(M,t) = M(t)^{-2} f(M/M(t)).
    Used throughout the theory (Supplement Eq. S17); paper reports collapse for selected parameters, but this is an input, not a derivation.
  • domain assumption Cluster radius R ∼ M^{1/d_f} with d_f = d = 2; clusters are compact.
    Supplement says 'Assuming compact, non-fractal clusters where the spatial dimension equals the fractal dimension'; fractality is acknowledged as emerging at high density and large J, so the simplification may fail in some regimes.
  • ad hoc to paper Power-law forms Vc ∼ M^γ and τp ∼ M^ν hold over the aggregation range.
    The authors state 'we will assume power-law dependencies in our theory below in order to capture the limiting values of z', making this an assumption fitted to the data rather than derived.
  • ad hoc to paper In the collective regime, clusters have equal scalar speeds but randomly oriented velocities, supplying a non-zero mass-independent relative speed in the ballistic kernel.
    Implied by using γ=0 in the kernel K ∼ |Vc(M) - Vc(M')| while obtaining z=2; the written kernel is zero if V is purely a scalar speed, so this premise is needed but unstated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Collective ballistic motion explains fast aggregation in adhesive active matter." pith.science (2026). https://pith.science/paper/XCEGH6IY

@misc{pith2026250811793,
  author       = {Pith},
  title        = {Pith review of: Collective ballistic motion explains fast aggregation in adhesive active matter},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XCEGH6IY}},
  note         = {Machine review of arXiv:2508.11793}
}
read the original abstract

Inspired by motile cells in tissue formation, we find that active systems of self-aligning adhesive particles undergo ballistic aggregation through a flocking transition. This kinetic regime emerges when the cluster persistence length grows faster with cluster mass than the intercluster distance does. We also identify and explain distinct non-collective kinetic regimes, including biologically relevant long-lived transients. Our analytical and numerical results offer a unified framework explaining the broad range of experimentally observed aggregation exponents in cellular systems and reveal physical principles potentially critical for timely tissue organization.

Figures

Figures reproduced from arXiv: 2508.11793 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. b summarizes the aggregation regimes in the Pe–J plane. At low J and Pe the system follows diffusion￾limited cluster aggregation (Region I); increasing Pe at 100 101 102 103 101 102 103 104 M t J = 0 0.1 1 10 100 (a) t t 2 t 1/2 10−1 100 101 102 103 P e 10−4 10−3 10−2 10−1 100 101 102 J 0.5 1 1.5 2 (b) z Collective ballistic III Diffusion-limited I Ballistic II FIG. 4. High persistence enables non-collective ballist… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 41 canonical work pages

  1. [1]

    Xiong and S

    C. Xiong and S. Friedlander, Proceedings of the National Academy of Sciences 98, 11851 (2001)

    work page 2001

  2. [2]

    A. A. Hyman, C. A. Weber, and F. Jülicher, Annual re- view of cell and developmental biology 30, 39 (2014)

    work page 2014

  3. [3]

    S. V. Franklin and M. D. Shattuck, Handbook of granular materials (CRC Press, 2016)

    work page 2016

  4. [4]

    de Castro, F

    P. de Castro, F. Urbina, A. Norambuena, and F. Guzmán-Lastra, Physical Review E 108, 044104 (2023)

    work page 2023

  5. [5]

    C. B. Caporusso, L. F. Cugliandolo, P. Digregorio, G. Gonnella, D. Levis, and A. Suma, Physical Review Letters 131, 068201 (2023)

    work page 2023

  6. [6]

    Grobas, M

    I. Grobas, M. Polin, and M. Asally, Elife 10, e62632 (2021)

    work page 2021

  7. [7]

    O’Toole, H

    G. O’Toole, H. B. Kaplan, and R. Kolter, Annual Re- views in Microbiology 54, 49 (2000)

    work page 2000

  8. [8]

    N. C. Heer and A. C. Martin, Development 144, 4249 (2017)

    work page 2017

Show all 44 references
  1. [9]

    S. G. Krens and C.-P. Heisenberg, Current topics in de- velopmental biology 95, 189 (2011)

    2011

  2. [10]

    E. F. Teixeira, C. P. Beatrici, H. C. Fernandes, and L. G. Brunnet, Physical Review Letters 134, 138401 (2025)

    2025

  3. [11]

    Bothe, E

    M. Bothe, E. Lardet, A. Poliakov, G. Pruess- ner, T. Bertrand, and I. Bordeu, arXiv preprint arXiv:2503.13714 (2025)

    2025 arXiv

  4. [12]

    Méhes, E

    E. Méhes, E. Mones, V. Nemeth, and T. Vicsek, PloS one 7, e31711 (2012)

    2012

  5. [13]

    Leyvraz, Physics Reports 383, 95 (2003)

    F. Leyvraz, Physics Reports 383, 95 (2003)

    2003

  6. [14]

    Nakajima and S

    A. Nakajima and S. Ishihara, New Journal of Physics 13, 033035 (2011)

    2011

  7. [15]

    Cremer and H

    P. Cremer and H. Löwen, Physical Review E 89, 022307 (2014)

    2014

  8. [16]

    Durand, Plos Computational Biology 17, e1008576 (2021)

    M. Durand, Plos Computational Biology 17, e1008576 (2021)

    2021

  9. [17]

    Jiang and F

    Y. Jiang and F. Leyvraz, Journal of Physics A: Mathe- matical and General 26, L179 (1993)

    1993

  10. [18]

    P. L. Krapivsky, S. Redner, and E. Ben-Naim, A kinetic view of statistical physics (Cambridge University Press, 2010)

    2010

  11. [19]

    Mones, A

    E. Mones, A. Czirók, and T. Vicsek, New journal of physics 17, 063013 (2015)

    2015

  12. [20]

    Baconnier, O

    P. Baconnier, O. Dauchot, V. Démery, G. Düring, S. Henkes, C. Huepe, and A. Shee, Reviews of Modern Physics 97, 015007 (2025)

    2025

  13. [21]

    J. M. Belmonte, G. L. Thomas, L. G. Brunnet, R. M. C. de Almeida, and H. Chaté, Phys. Rev. Lett. 100, 248702 (2008)

    2008

  14. [22]

    T. Paul, N. Vadakkayil, and S. K. Das, Physical Review E 109, 064607 (2024)

    2024

  15. [23]

    C. P. Beatrici, R. M. de Almeida, and L. G. Brunnet, Physical Review E 95, 032402 (2017)

    2017

  16. [24]

    M. v. Smoluchowski, Z. Physik. 17, 585 (1916)

    1916

  17. [25]

    Szabo, G

    B. Szabo, G. Szöllösi, B. Gönci, Z. Jurányi, D. Selmeczi, and T. Vicsek, Physical Review E 74, 061908 (2006)

    2006

  18. [26]

    Vicsek, A

    T. Vicsek, A. Czirók, E. Ben-Jacob, I. Cohen, and O. Shochet, Physical review letters 75, 1226 (1995)

    1995

  19. [27]

    Henkes, Y

    S. Henkes, Y. Fily, and M. C. Marchetti, Physical Review E—Statistical, Nonlinear, and Soft Matter Physics 84, 040301 (2011)

    2011

  20. [28]

    Sarkar, G

    D. Sarkar, G. Gompper, and J. Elgeti, Communications Physics 4, 36 (2021)

    2021

  21. [29]

    D. L. Barton, S. Henkes, C. J. Weijer, and R. Sknepnek, PLoS computational biology 13, e1005569 (2017)

    2017

  22. [30]

    Martín-Gómez, D

    A. Martín-Gómez, D. Levis, A. Díaz-Guilera, and I. Pag- onabarraga, Soft matter 14, 2610 (2018)

    2018

  23. [31]

    Y. Zhao, T. Ihle, Z. Han, C. Huepe, and P. Romanczuk, Physical Review E 104, 044605 (2021)

    2021

  24. [32]

    [ 13, 17, 18, 23, 24, 40–42]

    See Supplemental Material at [URL will be inserted by publisher] for videos and analytical details, which in- cludes Refs. [ 13, 17, 18, 23, 24, 40–42]

  25. [33]

    Peruani, J

    F. Peruani, J. Starruß, V. Jakovljevic, L. Søgaard- Andersen, A. Deutsch, and M. Bär, Physical review let- ters 108, 098102 (2012)

    2012

  26. [34]

    Fily and M

    Y. Fily and M. C. Marchetti, Physical review letters 108, 235702 (2012)

    2012

  27. [35]

    Y. Fily, S. Henkes, and M. C. Marchetti, Soft matter 10, 2132 (2014)

    2014

  28. [36]

    M. C. Marchetti, Y. Fily, S. Henkes, A. Patch, and D. Yl- lanes, Current Opinion in Colloid & Interface Science 21, 34 (2016)

    2016

  29. [37]

    Villa-Torrealba, C

    A. Villa-Torrealba, C. Chávez-Raby, P. de Castro, and R. Soto, Physical Review E 101, 062607 (2020)

    2020

  30. [38]

    de Castro, S

    P. de Castro, S. Diles, R. Soto, and P. Sollich, Soft Matter 17, 2050 (2021)

    2021

  31. [39]

    Kolb, Physical review letters 53, 1653 (1984)

    M. Kolb, Physical review letters 53, 1653 (1984)

    1984

  32. [40]

    Moncho-Jordá, F

    A. Moncho-Jordá, F. Martínez-López, M. Quesada- Pérez, M. Cabrerizo-Vílchez, and R. Hidalgo-Álvarez, Surface and Colloid Science , 113 (2004)

    2004

  33. [41]

    Moncho-Jordá, F

    A. Moncho-Jordá, F. Martınez-López, and R. Hidalgo- Álvarez, Physica A: Statistical Mechanics and its Appli- cations 282, 50 (2000)

    2000

  34. [42]

    Moncho-Jordá, G

    A. Moncho-Jordá, G. Odriozola, F. Martınez-López, A. Schmitt, and R. Hidalgo-Álvarez, The European Phys- ical Journal E 5, 471 (2001)

    2001

  35. [43]

    Cochet-Escartin, T

    O. Cochet-Escartin, T. T. Locke, W. H. Shi, R. E. Steele, and E.-M. S. Collins, Biophysical journal 113, 2827 (2017)

    2017

  36. [44]

    D. A. Beysens, G. Forgacs, and J. A. Glazier, Proceedings of the National Academy of Sciences 97, 9467 (2000). 1 Supplementary Material: Collective ballistic motion explains fast aggregation in adhesive active matter ADDITIONAL SIMULA TION DET AILS We use kc = 200 v0/(σµ), kad...

    2000

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.