Avalanche criticality emerges by thermal fluctuation in a quiescent glass

arXiv:2409.15775 · cond-mat.soft, cond-mat.dis-nn, cond-mat.stat-mech · Submitted 2024-09-24 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Avalanche criticality emerges by thermal fluctuation in a quiescent glass".

Mira: Here's the gist of the paper. Avalanche criticality emerges by thermal fluctuation in a quiescent glass > The paper reports avalanche criticality of thermal relaxation in glassy systems after a rapid quench by molecular simulation > Critical Phenomenon Analogy The study investigates whether avalanche criticality emerges in supercooled liquids and glasses driven solely by thermal fluctuation > The authors use MD simulations to monitor particle rearrangement events in quiescent glasses prepared by rapid quenching and subject them to small thermal fluctuations > They demonstrate that the rearrangement events consist of several local rearrangements linked by the elastic field > The size of these rearrangement events follows a power-law distribution, suggesting that particle rearrangement driven by thermal fluctuation is a critical avalanche dynamics > This observation provides firm ground to investigate the relation between intermittent dynamics and arXiv:2409.15775v1

cond-mat.soft: 24 Sep 2024 > Model Construction and Event Definition The model considered is a 50:50 mixture of two types of soft sphere particles in two dimensions, with particle sizes chosen to avoid crystallization > The interaction potential is an inverse-power-law potential defined by the equation v(rij) = ε

(σij/rij)12 − (σij/rc,ij)12 + C ((rij − rc,ij)/σij) + D ((rij − rc,ij)/σij)2: θ(rc,ij − rij), where parameters are defined > Rearrangement events are detected by analyzing the potential energy landscape using FIRE2.0 to obtain inherent structures (ISs) from time series of configurations > The size of rearrangements, S(t), is defined as the TSD of the particles between ISs at time t and t + δt, where S(t) = P i(ri,IS(t) − ri,IS(t + δt)) squared > Statistical Validation and Scaling A large set of simulations was performed to calculate the probability distribution P(S, N) of the TSD S for various system sizes N > A scaling analysis using the relation P(S, N) ∼ S−τ f(S/Sc) = S−τ c g(S/Sc), where Sc is an N-dependent cutoff size, was performed to ascertain if the power-law distribution originates from a critical phenomenon > The numerical data followed a power-law relationship consistent with Eq.

Kai: (2) for 1 Elastoplastic Interpretation The criticality is explained from the elastoplastic viewpoint, where thermal fluctuation causes a crossing of an energy barrier leading to local rearrangement > The elastic field links multiple local rearrangements, which constitute the avalanche, as evidenced by the quadrupolar character of displacement fields around local rearrangement regions > This suggests that avalanche dynamics disappear if aging dynamics after a finite waiting time is focused on, as intermittent rearrangements during aging are not system-spanning nor critical > The exponent τ ≃ 1 found in this work is significantly smaller than those observed in the stationary state of the thermal EPM, which is τ ≃ 1.3 > The study concludes that particle rearrangements can be understood as a critical phenomenon analogous to the fluidization of amorphous solids > The authors suggest that aging dynamics in glasses can be understood by considering glassy systems as “solids that flow” > Future work should investigate the waiting time and maturity dependence of rearrangement, potentially using thermal EPMs to understand aging in real structural glasses > The potential energy landscape analysis is versatile for measuring rearrangements in various situations, including equilibrium relaxation dynamics > The results hint at a new characterization of the difference between equilibrium/stationary state dynamics and aging dynamics of glasses based on the avalanche critical exponent > Supplemental Analysis Details The analysis confirms that the definition of S depends on the time interval δt, and reducing δt sufficiently small is expected to converge physical quantities > The scaling relation is also confirmed for dynamics with the Nos´e–Hoover thermostat at T = 0.1 with a time interval of δt = 0.1 > The analysis using the number of defects, Ndefect, confirms that the exponent of the power-law region is also τ ≃ 1 for this distribution > This indicates a robust finding across different simulation parameters and observation times > The analysis at lower temperatures, T = 0.02, also confirms the power-law relation with τ ≃ 1 > Definition of Avalanche Size The size of particle rearrangements is defined by the total squared displacements (TSD) S(t), which is calculated as S(t) = P i(ri,IS(t) − ri,IS(t + δt)) squared > Alternatively, the size can be defined by the number of defects Ndefect, where Ndefect = P iϕi and this definition is shown to not affect the result or discussion in the main text > The analysis using bond-breakage and bond-insertion processes also confirms that the exponent of the power-law region is τ ≃ 1 for Ndefect > The choice of definition does not affect the result or discussion in the main text > Error Evaluation Methodology The errors for moments and cutoffs are evaluated by the bootstrap method, where values are resampled Nbootstrap times to obtain new sets of S > Standard deviations computed from these distributions were used as the error for each value > This procedure ensures rigorous statistical evaluation of the derived quantities > The scaling analysis confirms that finite size scaling collapses the distribution on a single curve, supporting the critical phenomenon hypothesis > The error bars in (b) and (d) represent bootstrap standard errors > Key Findings on Exponents The calculated exponent τ ≃ 1 is significantly smaller than the ones observed in the stationary state of the thermal EPM, which is τ ≃ 1.3 > The steady shear flow of amorphous solids is also characterized by larger exponents: τ ≃ 1.36 in the two-dimensional EPM and τ ≃ 1.49 in the MD simulations with an athermal quasistatic protocol > These observations hint at a new characterization of the difference between equilibrium/stationary state dynamics and aging dynamics of glasses based on the avalanche critical exponent > The scaling relation is confirmed for Ndefect, showing that the exponent of the power-law region is also τ ≃ 1 > Reference List Summary

1: D.

Mira: S.

Kai: Fisher, K.

Paper summary: Kai: So, looking at the full picture of this paper on "Avalanche criticality emerges by thermal fluctuation in a quiescent glass," it really boils down to the idea that we can see avalanche behavior—that power-law distribution of particle rearrangements—arising purely from thermal fluctuation when you quench a glass rapidly one <ref:2409.15775#pg1,Avalanche criticality emerges by thermal fluctuation in a quiescent glass>.

Mira: The authors show that by modeling the system as an elastoplastic material driven by thermal fluctuation, they can explain dynamics below the glass transition point, which is important because it connects the intermittent rearrangement events to known critical phenomena in sheared amorphous solids six <ref:2409.15775#pg3,in sheared amorphous solids 6>.

Lev: The main implication for us is that we have a new way to characterize aging dynamics in glasses by looking at this avalanche critical exponent tau. It suggests that the scaling properties of these relaxation events are distinct from those observed in stationary thermal processes Key Findings on Exponents.

Kai: So, if you're listening and just want the simple version, this paper says that particle rearrangements in a glass can be understood as a critical phenomenon, analogous to the fluidization of amorphous solids three <ref:2409.15775#pg1>. It's about how thermal noise drives flow.

Mira: And the authors propose that aging dynamics in glasses should be understood by considering these systems as "solids that flow," which is a conceptual shift from thinking about them purely as static structures nine.

Lev: The practical hurdle, though, is that they are working with an idealized system of soft spheres and they are looking at quenched conditions, so we have to keep in mind that applying this directly to real structural glasses will require careful consideration of those differences <ref:2409.15775#pg2>.

Kai: That's the limitation—the model is simplified, but it gives us a framework for how thermal noise drives complex collective behaviors in disordered media fourteen.

Mira: And they suggest future work should focus on investigating the waiting time and maturity dependence of these rearrangements, potentially using thermal energy processes to understand aging in real structural glasses twenty-six <ref:2409.15775#pg1>. That's where the next steps lie.

Conclusion: Kai: So we've seen how thermal noise causes these particle rearrangements in a glass to look like an avalanche, and now we're looking at the final words on this paper, "Avalanche criticality emerges by thermal fluctuation in a quiescent glass."

Mira: Yeah, what they're really saying is that this power-law behavior you see isn't just some random noise; it’s tied to a specific critical point in how these glassy systems relax.

Lev: From an error correction standpoint, this suggests that the dynamics we see aren't just messy relaxation but follow a statistical rule, which would be helpful for designing better models for real hardware errors.

Kai: I think what they’re pointing to is that these avalanches are like a fluidization process in amorphous solids; it takes something solid and drives it toward flow through thermal energy.

Mira: Exactly, and the authors argue this gives us a way to describe aging dynamics in glasses—that slow, time-dependent behavior after you quench the system—by calling them "solids that flow."

Lev: If this holds up under real hardware constraints, it means we can predict how these systems will behave when they're being stressed or aged over long times.

Kai: It really shifts the focus from just seeing disorder to understanding the collective, critical movement within that disorder.

Mira: And what they do to show this is use a specific mathematical scaling relation that confirms this power-law connection across different system sizes.

Lev: The caveat here is how well the simulation captures those long-range elastic fields and if we can truly map those local rearrangements onto physical defects in a circuit.

Kai: So, it’s a nice piece of work because it provides a new language—a critical exponent—to separate what's happening in steady state versus what's happening while the glass is aging.

Mira: That separation is key; it suggests that the way we measure relaxation changes depending on whether you are looking at the immediate aftermath or the long-term evolution of a material.

Lev: We need to see if those exponents match up when we try to run this on actual simulated annealing hardware where we have real thermal noise sources.

Kai: Yeah, so the big question now is how this critical behavior translates from these simplified sphere models to the complex materials that actually make up our devices.

Yuki Takaha, *Hideyuki Mizuno, Atsushi Ikeda

Graduate School of Arts and Sciences, The University of Tokyo · Research Center for Complex Systems Biology, Universal Biology Institute, The University of Tokyo

cond-mat.soft, cond-mat.dis-nn, cond-mat.stat-mech

Submitted: 2024-09-24

Updated: 2024-09-24

Comments: 6 pages, 4 figures + 5 pages, 6 figures in Supplemental Material

DOI: 10.1103/syhs-c28g

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 95/100

The gist: Avalanche criticality emerges by thermal fluctuation in a quiescent glass > The paper reports avalanche criticality of thermal relaxation in glassy systems after a rapid quench by molecular…

Key concepts

Avalanche Criticality
This refers to a phenomenon where small thermal fluctuations can trigger a cascade of linked particle rearrangements. The size of these rearrangement events follows a power-law distribution, which is the hallmark of criticality, indicating that the system exhibits scale-free behavior when driven by thermal noise.
Molecular Dynamics (MD) Simulations
These are computer simulations used to model the behavior of atoms and molecules in a glass. The researchers used MD to observe particle movements and track how they rearrange over time when subjected to small thermal fluctuations, allowing them to study the underlying dynamics.
Elastoplastic Interpretation
The critical behavior is explained by viewing thermal fluctuations as events that cause particles to overcome energy barriers, leading to local rearrangements. These local events are connected by an elastic field, forming an 'avalanche' of coupled movements that define the critical nature of the dynamics.

Terminology

Summary

Avalanche criticality emerges by thermal fluctuation in a quiescent glass > The paper reports avalanche criticality of thermal relaxation in glassy systems after a rapid quench by molecular simulation >

Critical Phenomenon Analogy

The study investigates whether avalanche criticality emerges in supercooled liquids and glasses driven solely by thermal fluctuation > The authors use MD simulations to monitor particle rearrangement events in quiescent glasses prepared by rapid quenching and subject them to small thermal fluctuations > They demonstrate that the rearrangement events consist of several local rearrangements linked by the elastic field > The size of these rearrangement events follows a power-law distribution, suggesting that particle rearrangement driven by thermal fluctuation is a critical avalanche dynamics > This observation provides firm ground to investigate the relation between intermittent dynamics and arXiv:2409.15775v1 [cond-mat. p. 0soft] 24 Sep 2024 >

Model Construction and Event Definition

The model considered is a 50:50 mixture of two types of soft sphere particles in two dimensions, with particle sizes chosen to avoid crystallization > The interaction potential is an inverse-power-law potential defined by the equation v(rij) = ε[(σij/rij)12 − (σij/rc,ij)12 + C ((rij − rc,ij)/σij) + D ((rij − rc,ij)/σij)2]θ(rc,ij − rij), where parameters are defined > Rearrangement events are detected by analyzing the potential energy landscape using FIRE2.0 to obtain inherent structures (ISs) from time series of configurations > The size of rearrangements, S(t), is defined as the TSD of the particles between ISs at time t and t + δt, where S(t) = P i(ri,IS(t) − ri,IS(t + δt)) squared >

Statistical Validation and Scaling

A large set of simulations was performed to calculate the probability distribution P(S, N) of the TSD S for various system sizes N > A scaling analysis using the relation P(S, N) ∼ S−τ f(S/Sc) = S−τ c g(S/Sc), where Sc is an N-dependent cutoff size, was performed to ascertain if the power-law distribution originates from a critical phenomenon > The numerical data followed a power-law relationship consistent with Eq. (2) for 1

Elastoplastic Interpretation

The criticality is explained from the elastoplastic viewpoint, where thermal fluctuation causes a crossing of an energy barrier leading to local rearrangement > The elastic field links multiple local rearrangements, which constitute the avalanche, as evidenced by the quadrupolar character of displacement fields around local rearrangement regions > This suggests that avalanche dynamics disappear if aging dynamics after a finite waiting time is focused on, as intermittent rearrangements during aging are not system-spanning nor critical > The exponent τ ≃ 1 found in this work is significantly smaller than those observed in the stationary state of the thermal EPM, which is τ ≃ 1.3 >

Conclusion and Future Directions

The study concludes that particle rearrangements can be understood as a critical phenomenon analogous to the fluidization of amorphous solids > The authors suggest that aging dynamics in glasses can be understood by considering glassy systems as “solids that flow” > Future work should investigate the waiting time and maturity dependence of rearrangement, potentially using thermal EPMs to understand aging in real structural glasses > The potential energy landscape analysis is versatile for measuring rearrangements in various situations, including equilibrium relaxation dynamics > The results hint at a new characterization of the difference between equilibrium/stationary state dynamics and aging dynamics of glasses based on the avalanche critical exponent >

Supplemental Analysis Details

The analysis confirms that the definition of S depends on the time interval δt, and reducing δt sufficiently small is expected to converge physical quantities > The scaling relation is also confirmed for dynamics with the Nos´e–Hoover thermostat at T = 0.1 with a time interval of δt = 0. p. 11 > The analysis using the number of defects, Ndefect, confirms that the exponent of the power-law region is also τ ≃ 1 for this distribution > This indicates a robust finding across different simulation parameters and observation times > The analysis at lower temperatures, T = 0. p. 702, also confirms the power-law relation with τ ≃ 1 >

Definition of Avalanche Size

The size of particle rearrangements is defined by the total squared displacements (TSD) S(t), which is calculated as S(t) = P i(ri,IS(t) − ri,IS(t + δt)) squared > Alternatively, the size can be defined by the number of defects Ndefect, where Ndefect = P iϕi and this definition is shown to not affect the result or discussion in the main text > The analysis using bond-breakage and bond-insertion processes also confirms that the exponent of the power-law region is τ ≃ 1 for Ndefect > The choice of definition does not affect the result or discussion in the main text >

Error Evaluation Methodology

The errors for moments and cutoffs are evaluated by the bootstrap method, where values are resampled Nbootstrap times to obtain new sets of S > Standard deviations computed from these distributions were used as the error for each value > This procedure ensures rigorous statistical evaluation of the derived quantities > The scaling analysis confirms that finite size scaling collapses the distribution on a single curve, supporting the critical phenomenon hypothesis > The error bars in (b) and (d) represent bootstrap standard errors >

Key Findings on Exponents

The calculated exponent τ ≃ 1 is significantly smaller than the ones observed in the stationary state of the thermal EPM, which is τ ≃ 1.3 > The steady shear flow of amorphous solids is also characterized by larger exponents: τ ≃ 1. p. 436 in the two-dimensional EPM and τ ≃ 1. p. 449 in the MD simulations with an athermal quasistatic protocol > These observations hint at a new characterization of the difference between equilibrium/stationary state dynamics and aging dynamics of glasses based on the avalanche critical exponent > The scaling relation is confirmed for Ndefect, showing that the exponent of the power-law region is also τ ≃ 1 >

Reference List Summary

[1] D. S. Fisher, K. Dahmen, S. Ramanathan, and Y. p. 4 BenZion, Statistics of earthquakes in simple models of heterogeneous faults, Physical review letters 78, 4885 (1997) >

[2] K. Dahmen, D. Erta¸s, and Y. p. 4 Ben-Zion, GutenbergRichter and characteristic earthquake behavior in simple mean-field models of heterogeneous faults, Physical Review E 58, 1494 (1998) >

[3] E. M. Kramer and A. p. 4 E. Lobkovsky, Universal power law in the noise from a crumpled elastic sheet, Physical Review E 53, 1465 (1996) >

[4] D. Shohat, Y. Friedman, and Y. p. 4 Lahini, Logarithmic aging via instability cascades in disordered systems, Nature Physics 19, 1890 (2023) >

[5] J. P. Sethna, K. Dahmen, S. Kartha, J. A. Krumhansl, B. p. 4 W. Roberts, and J. p. 4 D. Shore, Hysteresis and hierarchies: Dynamics of disorder-driven first-order phase transformations, Physical Review Letters 70, 3347 (1993) >

[6] C. E. Maloney and A. p. 4 Lemaitre, Amorphous systems in athermal, quasistatic shear, Physical Review E 74, 016118 (2006) >

[7] J. P. Sethna, K. A. Dahmen, and C. p. 4 R. Myers, Crackling noise, Nature 410, 242 (2001) >

[8] N. Oyama, H. Mizuno, and A. p. 0 p. 0 Ikeda, Unified view of avalanche criticality in sheared glasses, Physical Review E 104, 015002 (2021) >

[9] A. Nicolas, E. p. 4 E. Ferrero, K. p. 4 Martens, and J. p. 4-L. Barrat, Deformation and flow of amorphous solids: Insights from elastoplastic models, Reviews of Modern Physics 90, 045006 (2018) >

[10] M. Talamali, V. p. 4 Pet¨aj¨a, D. p. 4 Vandembroucq, and S. p. 4 Roux, Avalanches, precursors, and finite-size fluctuations in a mesoscopic model of amorphous plasticity, Physical Review E 84, 016115 (2011) >

[11] S. Karmakar, E. p. 4 Lerner, and I. p. 4 Procaccia, Statistical physics of the yielding transition in amorphous solids, Physical Review E 82, 055103 (2010) >

[12] J. Lin and M. Wyart, Mean-Field Description of Plastic Flow in Amorphous Solids, Physical review X 6, 011005 (2016) >

[13] L. Berthier and G. p. 4 Biroli, Theoretical perspective on the glass transition and amorphous materials, Reviews of modern physics 83, 587 (2011) >

[14] M. Ozawa and G. p. 4 Biroli, Elasticity, Facilitation, and Dynamic Heterogeneity in Glass-Forming Liquids, Physical Review Letters 130, 138201 (2023) >

[15] A. Tahaei, G. Biroli, M. Ozawa, M. Popovi´c, and M. p. 4 Wyart, Scaling Description of Dynamical Heterogeneity and Avalanches of Relaxation in Glass-Forming Liquids, Physical Review X 13, 031034 (2023) >

[16] L. Buisson, L. p. 4 Bellon, and S. p. 4 Ciliberto, Intermittency in ageing, Journal of Physics: Condensed Matter 15, S1163 (2003) >

[17] D. El Masri, L. p. 4 Berthier, and L. p. 4 Cipelletti, Subdiffusion and intermittent dynamic fluctuations in the aging regime of concentrated hard spheres, Physical Review E 82, 031503 (2010) >

[18] T. Yanagishima, J. p. 5 Russo, and H. p. 5 Tanaka, Common mechanism of thermodynamic and mechanical origin for ageing and crystallization of glasses, Nature communications 8, 15954 (2017) >

[19] E. Sanz, C. Valeriani, E. p. 5 Zaccarelli, W. p. 5 C. K. Poon, M. E. Cates, and P. p. 5 N. Pusey, Avalanches mediate crystallization in a hard-sphere glass, Proceedings of the National Academy of Sciences 111, 75 (2014) >

[20] V. Testard, L. p. 5 Berthier, and W. p. 4 Kob, Intermittent dynamics and logarithmic domain growth during the spinodal decomposition of a glass-forming liquid, The Journal of chemical physics 140, 164502 (2014) >

[21] H. Mizuno, H. Tong, A. Ikeda, and S. p. 4 Mossa, Intermittent rearrangements accompanying thermal fluctuations distinguish glasses from crystals, The Journal of Chemical Physics 153, 154501 (2020) >

[22] J. Gu´enol´e, W. p. 5 G. N¨ohring, A. p. 5 Vaid, F. Houll´e, Z. p. 5 Xie, A. Prakash, and E. p. 5 Bitzek, Assessment and optimization of the fast inertial relaxation engine (FIRE) for energy minimization in atomistic simulations and its implementation in LAMMPS, Computational Materials Science 175, 109584 (2020) >

[23] N. Oyama, H. Mizuno, and A. Ikeda, Shear-induced criticality in glasses shares qualitative similarities with the Gardner phase, Soft Matter 19, 6074 (2023) >

[24] Y. Nishikawa, M. p. 5 Ozawa, A. Ikeda, P. Chaudhuri, and L. p. 5 Berthier, Relaxation dynamics in the energy landscape of glass-forming liquids, Physical Review X 12, 021001 (2022) >

[25] In larger systems, the rearrangements occur more frequently, and then we need to use smaller time intervals δt to isolate individual rearrangement events. This requires huge numerical costs >

[26] J. Lin, E. Lerner, A. Rosso, and M. p. 5 Wyart, Scaling description of the yielding transition in soft amorphous solids at zero temperature, Proceedings of the National Academy of Sciences 111, 14382 (2014) >

[27] S. Papanikolaou, Shearing a glass and the role of pinning delay in models of interface depinning, Physical Review E 93, 032610 (2016) >

[28] E. A. Jagla, Avalanche-size distributions in mean-field plastic yielding models, Physical Review E 92, 042135 (2015) >

[29] S. Franz and S. p. 5 Spigler, Mean-field avalanches in jammed spheres, Physical Review E 95, 022139 (2017) >

[30] B. Shang, P. Guan, and J. p. 5-L. Barrat, Elastic avalanches reveal marginal behavior in amorphous solids, Proceedings of the National Academy of Sciences 117, 86 (2020) >

[31] J. C. Dyre, Colloquium: The glass transition and elastic models of glass-forming liquids, Reviews of modern physics 78, 953 (2006) >

[32] A. Lemaˆıtre, Structural Relaxation is a Scale-Free Process, Physical review letters 113, 245702 (2014) >

[33] L. F. Cugliandolo and J. p. 5 p. 5 Kurchan, Analytical solution of the off-equilibrium dynamics of a long-range spin-glass model, Physical Review Letters 71, 173 (1993) >

[34] L. F. Cugliandolo and J. Kurchan, On the out-ofequilibrium relaxation of the sherrington-kirkpatrick model, Journal of Physics A: Mathematical and General 27, 5749 (1994) >

[35] T. Odagaki and Y. p. 5 Hiwatari, Stochastic model for the glass transition of simple classical liquids, Physical Review A 41, 929 (1990) >

[36] C. Monthus and J. p. 5-P.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved understanding of aging dynamics in glasses: The AI system can model aging by considering glassy systems as solids that flow, which involves tracking rare local particle rearrangements and their resulting elastic dynamic facilitation even in aging glasses.

  2. Enhanced prediction of critical behavior: The system can predict the avalanche criticality by analyzing the scaling relation P(S, N) ∼ S-τ f(S/Sc) = S-τ c g(S/Sc), where the exponent tau is found to be approximately 1, allowing for more accurate characterization of critical phenomena in thermal fluctuations.

  3. Development of a new model for supercooled liquids: The AI can utilize the thermal EPM framework to study dynamics driven solely by thermal fluctuation, providing a coarse-grained model that reveals dynamically correlated regions emerge by stress relaxation triggered by thermal fluctuation and the dynamic facilitation by elastic fields.

  4. Improved detection of critical events: The system can detect critical avalanche dynamics in quiescent glasses after a rapid quench by monitoring particle rearrangement events using potential energy landscape analysis, specifically identifying local rearrangements linked by the elastic field that constitute a system-spanning event.

  5. Quantitative analysis of event size: The AI can quantitatively determine the size distribution of rearrangement events using TSD as an indicator, confirming that the size of the rearrangement events follows a power-law distribution, which is a hallmark of critical avalanche dynamics.

Related papers