Pith. sign in

REVIEW 4 major objections 6 minor 38 references

Coarse grained descriptions of the dynamics of yielding of amorphous solids under cyclic shear

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Adding mechanical noise feedback to the Ehrenfest model makes cyclic shear yielding a genuine dynamical transition, with a diverging timescale and a universal scaling form.

desk verdict A clean mean-field feedback model for cyclic shear yields a transition and a tan^2 scaling, but the key linearization is not checked for N-sensitivity, so the quantitative claims are conditional. read the letter →

arxiv 2505.14912 v1 pith:J77TLJN5 submitted 2025-05-20 cond-mat.stat-mech

Debargha Sarkar , Jishnu N. Nampoothiri , Muhittin Mungan , Jack T. Parley , Peter Sollich , Srikanth Sastry This is my paper · ORCID
classification cond-mat.stat-mech
keywords cyclicshearamorphoussolidsyieldingEhrenfestmodelmechanicalnoisedynamicaltransitionfatiguefailurecoarsegraining
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

This paper claims that the puzzling many-cycle behaviour of amorphous solids under cyclic shear (the divergence of the number of cycles to steady state, non-monotonic fatigue-failure dynamics, and coexisting frozen and fluidized states) can be explained by a coarse-grained description in which mechanical noise feeds back through the fraction of unstable blocks, $D(t)=\alpha P_U(t)$. The feedback turns the Ehrenfest-type random-walk model into one with a genuine dynamical transition at a critical strain amplitude $\gamma^*$, below which only a frozen state with $P_U=0$ exists and above which a fluidized state appears, with the escape time from stable states diverging as $(\gamma-\gamma^*)^{-1}$. A minimal three-state coarse graining (unstable, threshold, absorbing) is shown to reproduce the non-monotonic fatigue-failure dynamics that a two-state description cannot capture. Near the annealing-dependent critical strain $\gamma_c$, the unstable fraction follows the universal scaling $P_U=\delta\gamma[1+\tan^2(a\sqrt{\delta\gamma}\,\delta t)]$. If correct, these results identify mechanical noise feedback, rather than thermal activation, as the mechanism that sets the yield point and controls the fatigue limit.

What carries the argument

The load-bearing object is the feedback relation $D(t)=\alpha P_U(t)$, which makes the transition rates out of stable states depend on the instantaneous fraction of unstable blocks. With this feedback the coarse-grained master equation becomes nonlinear, and a fixed-point analysis yields the dynamical transition and the exponent $-1$ divergence of the escape time. A slow-manifold reduction near the critical frozen state produces the $\tan^2$ scaling (19). The three-state coarse graining (unstable, threshold, absorbing) is the minimal description that retains the non-monotonic fatigue-failure dynamics, which the two-state description cannot show.

What would settle it

Measure in atomistic simulations of cyclic shear the instantaneous strain diffusion coefficient $D(t)$ and the unstable fraction $P_U(t)$ near the yield amplitude; if $D/P_U$ is not approximately constant, or if the number of cycles to steady state does not diverge as $(\gamma-\gamma^*)^{-1}$, the transition mechanism does not hold. The universal scaling $P_U/\delta\gamma = 1+\tan^2(a\sqrt{\delta\gamma}\,\delta t)$ is directly testable: a failure to collapse data onto this curve near $\gamma_c$ would falsify the claimed universality.

Watch

Extended reading notes

Core claim

The central claim is that a self-consistent mechanical-noise feedback term converts the previously studied Ehrenfest model of cyclic shear yielding from a model with a crossover into one with a true dynamical phase transition. The transition is characterized by a critical strain amplitude $\gamma^*$: for $\gamma<\gamma^*$ the only steady state has $P_U=0$ (frozen), while for $\gamma>\gamma^*$ a fluidized fixed point with $P_U\neq 0$ emerges, and the mean escape time from stable states diverges as $(\gamma-\gamma^*)^{-1}$ approaching the transition from above. The paper further claims that a three-state coarse graining (unstable, threshold, absorbing) is the minimal description that shows non-monotonic 'fatigue failure' dynamics (plastic activity first drops, then rises to failure), and that near the critical strain the unstable fraction follows the universal form $P_U=\delta\gamma[1+\tan^2(a\sqrt{\delta\gamma}\,\delta t)]$. These results are presented as explaining the divergence of cycles, the annealing-dependent yield strain, and the coexistence of frozen and fluidized states seen in simulations.

Load-bearing premise

The whole argument rests on the assumption that the mechanical noise a stable region feels is strictly proportional to how many regions are currently unstable, so that escape rates vanish linearly as the unstable population shrinks.

Editorial extensions

If this is right

  • The number of cycles to reach a steady state diverges as $(\gamma-\gamma^*)^{-1}$ on either side of the yield amplitude, matching the fatigue-limit behaviour reported in simulations.
  • Below $\gamma^*$ the system always freezes, but different initial annealing states lead to different frozen configurations; above $\gamma^*$ a fluidized state coexists with frozen attractors whose basin shrinks as $\gamma$ grows.
  • The critical strain $\gamma_c$ for a well-annealed sample is larger the better the sample is annealed, consistent with the observation that well-annealed glasses yield at higher amplitudes.
  • Near $\gamma_c$, the relaxation time to reach the fatigue minimum and the time to re-fluidize both scale as $\delta\gamma^{-1/2}$, following from the $\tan^2$ form.
  • A two-state description cannot produce non-monotonic fatigue-failure dynamics; three state classes (unstable, threshold, absorbing) are necessary, with threshold-state accumulation preceding failure.

Reading between the lines

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

  • If the linear feedback assumption $D=\alpha P_U$ fails in real systems (for instance because Eshelby stress fluctuations from distant events are delayed or nonlocal), the exponent $-1$ and the $\tan^2$ form would likely change, but a qualitatively similar frozen-fluidized transition might persist; measuring $D(t)$ against $P_U(t)$ in simulations would distinguish these cases.
  • The $\tan^2$ scaling is derived for a broad class of master equations, so it may apply to other driven disordered systems where activity feeds back into local stability, not just cyclic shear of glasses.
  • The three-state picture suggests that fatigue failure is preceded by an accumulation of probability in threshold (marginally stable) states; this could be probed in simulations by tracking the population of such states as a function of cycle number.
  • The same feedback mechanism could be incorporated into continuum elasto-plastic models, where making the local yield stress depend on instantaneous plastic activity might reproduce the fatigue-limit divergence without invoking thermal activation.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper studies a mean-field model of amorphous solids under cyclic shear, generalizing the Ehrenfest model of Mungan and Sastry by adding self-consistent mechanical noise D(t)=alpha P_U(t). In a two-state coarse graining, the authors find a dynamical transition at a critical strain amplitude gamma*: below gamma* only the frozen state P_U=0 exists, above it a fluidized state appears, and the escape time diverges as (gamma - gamma*)^(-1) (Eq. 9). They introduce a three-state model (unstable, threshold, absorbing) and show that it yields non-monotonic fatigue-failure trajectories and a basin of attraction that shrinks with increasing gamma, so that well-annealed samples have a larger critical strain gamma_c. Near gamma_c they propose a universal scaling P_U = delta_gamma [1 + tan^2(a sqrt(delta_gamma) delta_t)] (Eq. 19). Numerical trajectories of the full and coarse-grained models are presented in Figs. 1-3 to support these claims.

Significance. If substantiated, the paper would provide a minimal mean-field explanation of several simulation observations: the divergence of cycles near yielding, the annealing-dependent yield point, the non-monotonic property changes, and the spectrum of frozen states. Its strengths are the explicit master equations, the transparent fixed-point analysis in the two-state model, and the numerical demonstrations in Figs. 1-3. However, the manuscript's central derivations are almost entirely deferred to an unavailable Supplemental Material, and the linearization underlying the exponent might be sensitive to the number of states N; these issues must be addressed before the claims can be accepted as established.

major comments (4)
  1. [Eq. (6), Eqs. (16-18), Eq. (19); Ref. [31]] The formulas for W_US in Eq. (6), the three-state coarse-grained rates in Eqs. (16)-(18), and the derivation of the universal scaling (19) are all referenced to Ref. [31], which is listed as 'Supplemental Material (To be included) (2025)' and is not part of the submitted manuscript. These are load-bearing derivations: without them, Eq. (6) is an assertion, and the claimed 'full generality' of Eq. (19) is unverifiable. Please include the SM in the submission or move the key steps to the main text.
  2. [Divergence of timescales; Eqs. (6), (10)-(11)] The linearization W_US = alpha Y W'_US and W_SS = alpha Y W'_SS is justified by the D->0 behavior of 1/tau_j in Eq. (3). For a stable state j > k, r_j(D) = 2D/[tau0((sqrt(eps_j)-gamma)^2 + 2D)], which is linear only for D << (sqrt(eps_j)-gamma)^2. Since the gap (sqrt(eps_j)-gamma) vanishes as j -> k, the number of states for which the linearization fails at a given small P_U grows with N. The exact derivative dW_US/dP_U at P_U=0 involves a sum over q_j/(sqrt(eps_j)-gamma)^2, which is expected to grow with N, whereas the closed-form Eq. (6) gives a coefficient q_k p_{k-1,k}/(tau0 a) that does not exhibit such growth. Unless a cancellation is demonstrated, the transition condition alpha W'_US(gamma*) = W_SU(gamma*) and the exponent -1 in Eq. (9) may be finite-N artifacts. Please provide the SM derivation of Eq. (6) and an N-scaling check, for example by varying N in simulations of the full master equation.
  3. [Introduction, first paragraph and footnote [18]] The paper claims a power-law divergence of cycles on both sides of the yielding point, but the only derived behavior is the tau_US ~ (gamma - gamma*)^(-1) divergence for gamma > gamma* (Eq. (9)). Footnote [18] acknowledges that elasto-plastic simulations instead indicate a logarithmic divergence when the yielding point is approached from below. The model's prediction for the approach from below is not presented, and the discrepancy is not discussed. Clarify which side the model addresses and how the logarithmic behavior might arise within the same framework.
  4. [Generalised Master Equation with mechanical noise; Eqs. (16)-(18)] The non-monotonic fatigue-failure dynamics of the three-state model is a built-in consequence of the assumption that stable-state escape rates are proportional to P_U: every term on the right-hand side of Eqs. (16)-(18) carries a factor P_U, so an initial decay of P_U is automatic unless the threshold population P_T grows. The statement that the model 'predicts' fatigue failure dynamics (page 4) should be tempered; the novel content is the existence of the critical frozen state, the basin-of-attraction dependence on annealing, and the scaling form (19). I suggest rephrasing to emphasize the consistency-check role of the three-state construction.
minor comments (6)
  1. [Eq. (7)] The steady-state relation P*_U = W_US/(W_US + W_SU) should be written with the argument of W_US made explicit, e.g., W_US(P*_U), to avoid an apparent algebraic definition of the fixed point.
  2. [Fig. 1(d)] The caption refers to tau_lin without defining it in the text; please define tau_lin, presumably the timescale obtained from linearizing Eq. (5).
  3. [Eq. (12) and text before it] The notation W'_US and W'_SS is introduced without explicit definitions; please provide the definitions that are claimed to be in the SM.
  4. [Just after Eq. (14)] The counting argument gives a solution space of dimension N_s - 2, which for N_s = 2 yields isolated solutions but does not by itself prove uniqueness of the critical frozen state; please clarify whether uniqueness has been checked.
  5. [Eq. (3)] The denominator (sqrt(eps_j)-gamma)^2 is formatted with a superscript that may be misread; please use consistent notation, e.g., (\sqrt{\epsilon_j}-\gamma)^2.
  6. [General formatting] The paper chooses a half-Gaussian reference distribution (footnote [30]) but does not discuss the sensitivity of the transition and scaling results to this choice; a sentence on robustness across distributions would strengthen the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the transition, timescale divergence, and tan^2 scaling are derived consequences of explicitly stated model assumptions, not restatements of the inputs; self-citations are contextual rather than load-bearing.

full rationale

The paper's central results are derived from an explicit model: mechanical noise D(t) = alpha P_U(t) in Eq. (2), waiting times tau_j in Eq. (3), and transition probabilities in Eq. (1). The two-state transition, the (gamma - gamma*)^(-1) divergence, and the tan^2 scaling are mathematical consequences of these assumptions, in particular of the linearized rates W_US = alpha Y W'_US and W_SS = alpha Y W'_SS introduced in Eqs. (10)-(11). These are not identities imposed by defining the outputs in terms of the inputs: the order parameter P_U vanishes at the transition, and the divergence exponent follows from the assumed linear dependence, which is a stated modeling hypothesis rather than a hidden circular step. The non-monotonic fatigue failure dynamics emerges only in the three-state (N_u=1, N_s=2) reduction and is explicitly absent from the two-state model, so it is a nontrivial property of the model class rather than a restatement of the feedback assumption. The paper does draw on prior work by the same authors ([24]-[26]) for the Ehrenfest model and coarse-graining framework, but it restates the model and derives the new results in the present manuscript; these self-citations are contextual and not load-bearing. The skeptic's concern that the linearization W_US proportional to P_U is non-uniform near threshold states where sqrt(eps_j) - gamma can be small is a robustness or correctness issue about whether the model assumptions hold for finite N, not evidence that a prediction reduces to its input by construction. No fitted parameter is renamed as a prediction: the constant a in the scaling collapse fits the non-universal amplitude while the functional form is derived.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper's load-bearing machinery is a set of modeling choices: a mean-field feedback D=alpha*P_U, a linear vanishing of stable-state escape rates, a local-equilibrium coarse-graining, and an annealing-dependent initial condition. The transition and scaling results are mathematical consequences of these choices; they are not derived from atomistic physics and are not benchmarked against independent data. The main free parameters (alpha, beta, lambda, a) are not pinned down by comparison with simulation or experiment.

free parameters (4)
  • alpha (mechanical noise coupling)
    Coupling constant in D(t)=alpha*P_U(t) (Eq. 2) and in W_US (Eq. 6). It sets the strength of the feedback and controls the value of gamma*; no value is fitted from data, so it remains a free model parameter.
  • beta (annealing parameter in initial condition) = 40 used in Figs. 2 and 3
    Initial condition P_j(0) proportional to rho(epsilon_j) e^(beta*epsilon_j) (Eq. 4) parametrizes degree of annealing; the value beta=40 is chosen for the well-annealed sample. It changes gamma_c, the sample-dependent yield point.
  • lambda (threshold-state width)
    Defines the threshold states as epsilon in (gamma^2, gamma^2+lambda) in the three-state coarse graining; its value can be used to optimize the coarse-graining and affects the rates, but is not fixed by external data.
  • a (scaling collapse constant) = chosen to fit collapse in Fig. 3
    In Eq. (19) and Fig. 3(b,d), a is 'chosen to fit the scaling function'; it is a fit parameter for the universal tan^2 collapse.
assumptions (5)
  • domain assumption Mean-field mechanical-noise feedback: the strain diffusion coefficient is D(t)=alpha*P_U(t).
    Eq. (2) states that noise from plastic activity elsewhere is proportional to the instantaneous fraction of unstable blocks. This is the mechanism that makes stable states yield and is the engine of the dynamical transition.
  • domain assumption Stable-state escape rates vanish linearly with P_U as D goes to zero.
    Before Eq. (10), the paper retains only the linear dependence of stable-block transition rates on P_U, writing W_US = alpha*Y*W'_US and W_SS = alpha*Y*W'_SS. The existence of the critical frozen state and the tan^2 scaling depend on this linearization.
  • domain assumption A local equilibrium approximation within each macrostate gives the coarse-grained rates in the three-state model.
    The rates in Eqs. (16)-(18) are asserted to follow from local equilibrium inside the U, T, A macrostates, but the computation is deferred to the Supplemental Material [31], so it acts as an unverified modeling assumption here.
  • standard math The energy landscape is described by discrete states with a half-Gaussian distribution and stability range sqrt(epsilon).
    The model uses the reference distribution from [24] and the assignment delta-gamma = sqrt(epsilon) from [26]. These are carried over as standard inputs for the energy-landscape picture, not derived in this paper.
  • domain assumption Initial condition P_j(0) proportional to rho(epsilon_j) e^(beta*epsilon_j) represents the level of annealing.
    Eq. (4) introduces beta as an annealing parameter; the form of this initial condition is chosen, not derived, and the classification of frozen versus fluidized asymptotic states depends on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Coarse grained descriptions of the dynamics of yielding of amorphous solids under cyclic shear." pith.science (2026). https://pith.science/paper/J77TLJN5

@misc{pith2026250514912,
  author       = {Pith},
  title        = {Pith review of: Coarse grained descriptions of the dynamics of yielding of amorphous solids under cyclic shear},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J77TLJN5}},
  note         = {Machine review of arXiv:2505.14912}
}
read the original abstract

Recent computer simulations reveal several intriguing features in the evolution of properties of amorphous solids subjected to repeated cyclic shear deformation. These include the divergence of the number of cycles to reach steady states as the yielding point is approached, a non-monotonic change of properties with cycles, and the possibility of a spectrum of frozen states. Theoretical attempts to capture these properties through simple models, including the Ehrenfest model describing a random walk in a confining potential, have met partial success. Here, we show that incorporating the influence of mechanical noise through a feedback term leads to a genuine dynamical transition with characteristics reflecting those of yielding. Coarse graining the dynamics into a small number of variables leads to new insights regarding the dynamics of yielding.

Figures

Figures reproduced from arXiv: 2505.14912 by the authors.

Figure 1
Figure 1. FIG. 1. (a) Schematic diagram showing the energy of a state, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. (a) Non-monotonic time-dependence of [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 35 canonical work pages

  1. [25]

    J. T. Parley, S. Sastry, and P. Sollich, Physical Review Letters128, 198001 (2022)

    work page 2022

  2. [31]

    Sarkar, J

    D. Sarkar, J. Nambuthiri, M. Mungan, J. Parley, P. Sol- lich, and S. Sastry, Supplemental Material (To be in- cluded) (2025)

    work page 2025

  3. [18]

    Elasto-plastic simulations [34, 38], however, appear to indicate a logarithmic divergence as the yielding point is approached from below

    work page

  4. [1]

    Leishangthem, A

    P. Leishangthem, A. D. S. Parmar, and S. Sastry, Nature Communications8, 14653 (2017)

    work page 2017

  5. [2]

    Ozawa, L

    M. Ozawa, L. Berthier, G. Biroli, A. Rosso, and G. Tar- jus, Proceedings of the National Academy of Sciences 115, 6656 (2018)

    work page 2018

  6. [3]

    Bhaumik, G

    H. Bhaumik, G. Foffi, and S. Sastry, Proceedings of the National Academy of Sciences118, e2100227118 (2021)

    work page 2021

  7. [4]

    Y. Shi, M. B. Katz, H. Li, and M. L. Falk, Physical Review Letters98, 185505 (2007), publisher: American Physical Society

    work page 2007

  8. [5]

    W.-T. Yeh, M. Ozawa, K. Miyazaki, T. Kawasaki, and L. Berthier, Physical Review Letters124, 225502 (2020)

    work page 2020

Show all 38 references
  1. [6]

    V. V. Vasisht, G. Roberts, and E. Del Gado, Physical Re- view E102, 010604 (2020), publisher: American Physical Society

    2020

  2. [7]

    Popovi´ c, T

    M. Popovi´ c, T. W. J. De Geus, and M. Wyart, Physical Review E98, 040901 (2018)

    2018

  3. [8]

    Talamali, V

    M. Talamali, V. Pet¨ aj¨ a, D. Vandembroucq, and S. Roux, Comptes Rendus M´ ecanique340, 275 (2012)

    2012

  4. [9]

    H. J. Barlow, J. O. Cochran, and S. M. Fielding, Physical Review Letters125, 168003 (2020), publisher: American Physical Society

    2020

  5. [10]

    Kumar, S

    D. Kumar, S. Patinet, C. E. Maloney, I. Regev, D. Van- dembroucq, and M. Mungan, J. Chem. Phys.157, 174504 (2022)

    2022

  6. [11]

    Fiocco, G

    D. Fiocco, G. Foffi, and S. Sastry, Physical Review E88, 020301 (2013)

    2013

  7. [12]

    Kawasaki and L

    T. Kawasaki and L. Berthier, Physical Review E94, 022615 (2016)

    2016

  8. [13]

    Regev, T

    I. Regev, T. Lookman, and C. Reichhardt, Physical Re- view E88, 062401 (2013)

    2013

  9. [14]

    N. V. Priezjev, Computational Materials Science150, 162 (2018)

    2018

  10. [15]

    N. V. Priezjev, Computational Materials Science226, 112230 (2023)

    2023

  11. [16]

    Kurotani and H

    Y. Kurotani and H. Tanaka, Communications Materials 3, 67 (2022)

    2022

  12. [17]

    Maity, H

    S. Maity, H. Bhaumik, S. Athani, and S. Sastry, arXiv preprint arXiv:2409.17384 (2024)

    2024 arXiv

  13. [19]

    F. Kun, M. H. Costa, R. N. C. Filho, J. S. Andrade, J. B. Soares, S. Zapperi, and H. J. Herrmann, Journal of Statistical Mechanics: Theory and Experiment2007, P02003 (2007)

    2007

  14. [20]

    F. Kun, H. A. Carmona, J. S. Andrade, and H. J. Her- rmann, Physical Review Letters100, 094301 (2008), pub- lisher: American Physical Society

    2008

  15. [21]

    Basquin, Proc Am Soc Test Mater10, 625 (1910)

    O. Basquin, Proc Am Soc Test Mater10, 625 (1910)

    1910

  16. [22]

    Bhaumik, G

    H. Bhaumik, G. Foffi, and S. Sastry, The Journal of Chemical Physics156, 064502 (2022)

    2022

  17. [23]

    Sollich, F

    P. Sollich, F. m. c. Lequeux, P. H´ ebraud, and M. E. Cates, Phys. Rev. Lett.78, 2020 (1997)

    1997

  18. [24]

    Sastry, Physical Review Letters126, 255501 (2021)

    S. Sastry, Physical Review Letters126, 255501 (2021)

    2021

  19. [26]

    Mungan and S

    M. Mungan and S. Sastry, Physical Review Letters127, 248002 (2021)

    2021

  20. [27]

    Nicolas, E

    A. Nicolas, E. E. Ferrero, K. Martens, and J.-L. Barrat, Reviews of Modern Physics90, 045006 (2018)

    2018

  21. [28]

    H. A. Lockwood, E. S. Carrington, and S. M. Fielding, Ultra-delayed material failure via shear banding after straining an amorphous material, 2305.02144

    arXiv

  22. [29]

    J. O. Cochran, G. L. Callaghan, M. J. Caven, and S. M. Fielding, Physical Review Letters132, 168202 (2024), publisher: American Physical Society

    2024

  23. [30]

    The reference distribution in [26] was binomial, whereas here we chose the distribution to be half-Gaussian, as in [24]

  24. [32]

    H´ ebraud and F

    P. H´ ebraud and F. Lequeux, Physical Review Letters81, 2934 (1998)

    1998

  25. [33]

    Agoritsas, E

    E. Agoritsas, E. Bertin, K. Martens, and J.-L. Barrat, The European Physical Journal E38, 71 (2015)

    2015

  26. [34]

    C. Liu, E. E. Ferrero, E. A. Jagla, K. Martens, A. Rosso, and L. Talon, The Journal of Chemical Physics156, 104902 (2022)

    2022

  27. [35]

    Khirallah, B

    K. Khirallah, B. Tyukodi, D. Vandembroucq, and C. E. Maloney, Physical Review Letters126, 218005 (2021)

    2021

  28. [36]

    Bocquet, A

    L. Bocquet, A. Colin, and A. Ajdari, Physical Review Letters103, 036001 (2009)

    2009

  29. [37]

    J. T. Parley, S. M. Fielding, and P. Sollich, Physics of Fluids32, 127104 (2020)

    2020

  30. [38]

    Khandare and S

    P. Khandare and S. Sastry, Preprint

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.