Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids

arXiv:2602.22198 · cond-mat.dis-nn, cond-mat.soft · Submitted 2026-02-25 · 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: We're looking at the paper: "Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids".

Mira: Here's the gist of the paper. Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids The emergence of correlated spatiotemporal rearrangements is revealed through dynamical heterogeneities quantified by persistence measurements and the four-point susceptibility χ4 as temperature is varied As temperature increases, avalanche statistics evolve from scale-free behavior with exponential cutoffs to regimes dominated by systemspanning runaway events A system-size-dependent critical temperature Tc(L) separates intermittent avalanche dynamics from thermally assisted flow, where self-sustained avalanches transiently fluidize the system The relation between avalanche size and duration resembles that in sheared systems, whereas the statistics of minimal distances to yielding reveal a temperature-driven reorganization of marginal stability absent in strictly driven overdamped dynamics Our results demonstrate that thermal activation alone can generate a finite-size-controlled instability scale in disordered elastic media Model Implementation and Dynamics The study implements a coarse-grained elastoplastic model (EPM) of disordered solids where local relaxation events are induced by thermal fluctuations through activation over local energy barriers ∆E ∝ (σY −σi)a The model includes both deterministic yielding when σi ≥ σY and stochastic yielding for blocks with σi < σY due to thermal activation A block yields deterministically if its stress overcomes in absolute value the threshold (σi ≥ σY), and every block with σi < σY has a finite probability per unit time of yielding due to thermal activation (two different kinds of activation rules defined in Secs.

Kai: II A and II B below) When a block yields we choose the relaxation to be instantaneous, σi → σi − δσi, following

14: The stress drop tries to bring σi to zero, subtracting or adding its previous value, but adding also some stochasticity by drawing z as a random number from an exponential distribution of the form p(z) = 1/z0 e-z/z0 with z0 = 1.0 The local yielding of block i also affects the stress value in all other blocks j ≠ i, through σj → σj + Gij δσi (1) Characterization of Dynamical Heterogeneities Dynamical heterogeneities are quantified through persistence measurements and the associated four-point susceptibility χ4 Persistence is defined as p(t) = 1/L2 Σ i pi(t), where pi(t) = 1 if site i has not yet undergone a plastic event up to time t, and pi(t) = 0 if the site has been activated at least once The susceptibility χ4(t) is defined as χ4(t) = L2 h p 2(t) - ⟨p(t)⟩2 i (10) The position of the peak of χ4(t) shifts towards longer times as temperature decreases, which is consistent with the behavior of τα At low and decreasing temperatures, as T → 0, also the maximum of the susceptibility χmax4 (t) increases roughly as a power-law χmax4 ∼ T-0.8 Avalanche Statistics and Criticality Using extremal dynamics protocol, the distribution of avalanche sizes P(S) exhibits power-law behavior P(S) ∼ S−τ for certain critical x0 value xc≈ 0.022, with a τ exponent consistent with the one observed by Korchinski et al.

17: (τ≃ 1.5) Above x0 = xc, P(S) displays a sudden overshoot with an increasing probability accumulation at S≃ N as x0 increases The distributions P(D) of avalanche durations D are shown in Fig.

Kai: 4(c), and they look very similar to the distributions of SA, yet again the exponent that one can estimate for the power-law region is different from the previous ones: P(D) ∝ D−1.89 System-Size Dependence and Crossover A central result is that Tc depends on system size and decreases as L increases, with Tc ∝ L −0.23 ∝ N −0.115 Our data suggest that Tc(L) → 0 as L → ∞, albeit with a slow decay The apparent drift of τ reflects the increasing contribution of long-lived, merging avalanches as the system approaches Tc(L) from below The results clarify that thermal activation alone can reorganize marginal stability and avalanche dynamics in amorphous solids, introducing a finite-size-controlled instability scale that is absent in strictly driven overdamped systems The vanishing temperature (T = 0+) interpretation of the extremal protocol does not really stand, at least not from the moment that a x0 ≠ 0 is used The subcritical regime T < Tc(L) exhibits temperature-dependent power-law statistics, revealing a richer structure than that captured by extremal dynamics alone Discussion on System Behavior The behavior of P(S) in this controlled temperature protocol is very similar to the one shown in the previous section IV, which allows us to argue that the ‘extremal dynamics’ presented in

13, 14: is, in fact, an effectively finite–temperature dynamics rather than a T = 0+ one The changes in the τ exponent in P(S) as temperature increases is less evident for small system sizes, where curves look similar to the ones produced by the extremal dynamics, with an exponent τ≃ 1.5 fairly preserved (see Fig.

Paper summary: Mira: So looking at the title "Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids," it’s clear they are focusing on how temperature acts as this controlling knob for the dynamics of these materials.

Kai: I think what resonates most is that thermal activation alone can reorganize marginal stability and avalanche dynamics in amorphous solids, introducing a finite-size controlled instability scale that isn't present when you have strictly driven overdamped systems >

Lev: From an engineering standpoint, this means if you're designing something where thermal noise or fluctuations are present—like a material under certain temperature conditions—you need to account for this size-dependent critical temperature Tc(L) because it dictates whether your system behaves intermittently or starts running away completely >

Mira: The authors also point out that the vanishing temperature, T equals zero plus, interpretation of the extremal protocol doesn't really hold up once they introduce a non-zero x naught value, which is a significant caveat about how we interpret those low-temperature limits >

Kai: It’s important to remember that their results show that for temperatures below Tc(L), you still get power-law statistics with exponential cutoffs, which reveals a richer structure than just what you might expect from simpler extremal dynamics protocols >

Lev: I wonder how much overhead we'd have to build into a real error-correction scheme to account for these finite-size effects and the crossover behavior they describe, because if this scale is important, it might be relevant for understanding noise in those systems >

Conclusion: Kai: So, we've been looking at this paper about thermal activation driving avalanches in amorphous solids, and now we're hitting the conclusion section to see what it actually means for us >

Mira: It boils down to this idea that temperature isn't just a passive thing; it actively controls the whole instability process >

Kai: Right. The authors are showing how thermal activation creates a finite-size controlled instability scale in these disordered materials, which is something you don't see in strictly driven systems >

Lev: From my side, I have to ask about that size dependence. If Tc depends on the system size and decreases as the system gets bigger, what does that imply for scaling up our error correction schemes? >

Mira: Well, it suggests that even in an infinite system, if your temperature isn't perfectly zero, you might still trigger those system-spanning avalanches >

Kai: Exactly. And they found a critical temperature Tc that drops as the length of the material increases, which means arbitrarily small but finite temperatures could destabilize the intermittent regime >

Lev: So they're saying we need to be careful about assuming a fixed low temperature when modeling these complex dynamics on real hardware, because the system size matters fundamentally >

Mira: That’s what it points toward. And this whole setup is showing how thermal activation reorganizes marginal stability in these materials, which is pretty significant for understanding their long-term behavior >

Kai: It's interesting because it clarifies that thermal noise alone can generate a scale of instability that isn't there when you only use external driving forces >

Mira: And this leads us to the next thing. They also looked at how avalanche size relates to duration, which is something we need to look at next >

Gieberth Rodriguez-Lopez, Ezequiel E. Ferrero

Instituto de Fía del Sur (IFISUR), Universidad Nacional del Sur (UNS) · Instituto de Nanociencia y Nanotecnología, CNEA–CONICET, Centro Atómico Bariloche

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

Submitted: 2026-02-25

Updated: 2026-08-06

Comments: 15 pages, 10 figures, to appear in Phys. Rev. E

Journal ref: Phys. Rev. E 114, 025426 (2026)

DOI: 10.1103/3hjk-qlbn

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

Importance score: 83/100

The gist: Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids The emergence of correlated spatiotemporal rearrangements is revealed through dynamical…

Key concepts

Dynamical Heterogeneities
These refer to the non-uniformity of activity within a material over time. They are measured using persistence and the four-point susceptibility ($\chi_4$), which reveals how correlated spatiotemporal rearrangements occur across different parts of the solid.
Avalanche Statistics
This describes how avalanche sizes ($P(S)$) and durations ($P(D)$) are distributed. The paper finds that for temperatures below a critical point, these follow power laws, but above it, runaway events dominate.
Critical Temperature $T_c(L)$
This is a system-size dependent temperature that separates intermittent avalanche dynamics from thermally assisted flow. It decreases as the system size ($L$) increases, meaning even small finite temperatures can trigger large-scale avalanches in an infinite material.

Terminology

Summary

Thermal activation drives a finite-size crossover from scale-free to runaway avalanches in amorphous solids The emergence of correlated spatiotemporal rearrangements is revealed through dynamical heterogeneities quantified by persistence measurements and the four-point susceptibility χ4 as temperature is varied As temperature increases, avalanche statistics evolve from scale-free behavior with exponential cutoffs to regimes dominated by systemspanning runaway events A system-size-dependent critical temperature Tc(L) separates intermittent avalanche dynamics from thermally assisted flow, where self-sustained avalanches transiently fluidize the system The relation between avalanche size and duration resembles that in sheared systems, whereas the statistics of minimal distances to yielding reveal a temperature-driven reorganization of marginal stability absent in strictly driven overdamped dynamics Our results demonstrate that thermal activation alone can generate a finite-size-controlled instability scale in disordered elastic media

Model Implementation and Dynamics

The study implements a coarse-grained elastoplastic model (EPM) of disordered solids where local relaxation events are induced by thermal fluctuations through activation over local energy barriers ∆E ∝ (σY −σi)a The model includes both deterministic yielding when σi ≥ σY and stochastic yielding for blocks with σi < σY due to thermal activation A block yields deterministically if its stress overcomes in absolute value the threshold (σi ≥ σY), and every block with σi < σY has a finite probability per unit time of yielding due to thermal activation (two different kinds of activation rules defined in Secs. II A and II B below) When a block yields we choose the relaxation to be instantaneous, σi → σi − δσi, following [14] The stress drop tries to bring σi to zero, subtracting or adding its previous value, but adding also some stochasticity by drawing z as a random number from an exponential distribution of the form p(z) = 1/z0 e-z/z0 with z0 = 1. p. 20 The local yielding of block i also affects the stress value in all other blocks j ≠ i, through σj → σj + Gij δσi (1)

Characterization of Dynamical Heterogeneities

Dynamical heterogeneities are quantified through persistence measurements and the associated four-point susceptibility χ4 Persistence is defined as p(t) = 1/L2 Σ i pi(t), where pi(t) = 1 if site i has not yet undergone a plastic event up to time t, and pi(t) = 0 if the site has been activated at least once The susceptibility χ4(t) is defined as χ4(t) = L2 h p 2(t) - ⟨p(t)⟩2 i (10) The position of the peak of χ4(t) shifts towards longer times as temperature decreases, which is consistent with the behavior of τα At low and decreasing temperatures, as T → 0, also the maximum of the susceptibility χmax4 (t) increases roughly as a power-law χmax4 ∼ T-0.8

Avalanche Statistics and Criticality

Using extremal dynamics protocol, the distribution of avalanche sizes P(S) exhibits power-law behavior P(S) ∼ S−τ for certain critical x0 value xc≈ 0.022, with a τ exponent consistent with the one observed by Korchinski et al. p. 5 [17] (τ≃ 1.5) Above x0 = xc, P(S) displays a sudden overshoot with an increasing probability accumulation at S≃ N as x0 increases The distributions P(D) of avalanche durations D are shown in Fig. p. 5 4(c), and they look very similar to the distributions of SA, yet again the exponent that one can estimate for the power-law region is different from the previous ones: P(D) ∝ D−1. p. 689

System-Size Dependence and Crossover

A central result is that Tc depends on system size and decreases as L increases, with Tc ∝ L −0.23 ∝ N −0.115 Our data suggest that Tc(L) → 0 as L → ∞, albeit with a slow decay The apparent drift of τ reflects the increasing contribution of long-lived, merging avalanches as the system approaches Tc(L) from below

Conclusion on Thermal Activation

The results clarify that thermal activation alone can reorganize marginal stability and avalanche dynamics in amorphous solids, introducing a finite-size-controlled instability scale that is absent in strictly driven overdamped systems The vanishing temperature (T = 0+) interpretation of the extremal protocol does not really stand, at least not from the moment that a x0 ≠ 0 is used The subcritical regime T < Tc(L) exhibits temperature-dependent power-law statistics, revealing a richer structure than that captured by extremal dynamics alone

Discussion on System Behavior

The behavior of P(S) in this controlled temperature protocol is very similar to the one shown in the previous section IV, which allows us to argue that the ‘extremal dynamics’ presented in [13, 14] is, in fact, an effectively finite–temperature dynamics rather than a T = 0+ one The changes in the τ exponent in P(S) as temperature increases is less evident for small system sizes, where curves look similar to the ones produced by the extremal dynamics, with an exponent τ≃ 1.5 fairly preserved (see Fig. p. 8 S1 in Supp. p. 7 Mat. [41] for L = 64) The apparent drift of τ therefore reflects the increasing contribution of long-lived, merging avalanches as the system approaches Tc(L) from below The mechanism of avalanche merging or co-occurrence does not imply an increased plastic activity

Critical Temperature Scaling

The critical temperature Tc(L) decreases with L, with the value of η around η≃ 0.23, so Tc ∝ N −0.115 This suggests that in an infinite system even arbitrarily small but finite temperatures may trigger system-spanning avalanches The critical temperature is more controlled by multi-nucleation and an increasing “frequency” of plastic events, similar to what is discussed in [15]

Finite-Size Scaling at Fixed Temperature

Good collapses can be obtained for both P(S) and P(SA) curves with df≃ 1.22 and df≃ 1.29 for S and SA, respectively Up to L = 512 τ suffers little changes, but as size increases further, it shows more variation, decreasing from τ ∼ 1. p. 85 to τ ∼ 1.35 A way to justify this is by recalling that having chosen a fixed temperature for the analysis, increasing L brings us closer to the critical temperature (since Tc(L) decreases with L, Fig. p. 8 7), and the frequency of longerlived avalanches increases The apparent drift of τ therefore reflects the increasing contribution of long-lived, merging avalanches as the system approaches Tc(L) from below

Avalanche Size-Duration Relations

For ⟨S⟩, we first observe ⟨S⟩ ∝ Dδ with δ≃ 1.57 The similarity of δ between sheared and thermally activated systems suggests that the ratio df /z is robust to the triggering mechanism For ⟨SA⟩ we observe a law ⟨SA⟩ ∝ Dδ with δ≃ 1.8, with only a weak bending, which does not even develop in a saturation at the largest durations The relation is ⟨SA⟩ ∼ ⟨S⟩ 1. p. 515

Local Distance to Yielding

We define the local distance to yielding at site i as xi = σYi − σi, where σi is the local stress and σYi the local threshold We focus on the minimum distance xmin = mini xi, evaluated immediately after an avalanche ends Around T = Tc we measure a rather large exponent, θ≃ 1.7, as shown in Fig. p. 6 10(b) This can be seen as a feature of thermal systems, since in driving yielding the situation is always the opposite one: the occurrence of a ‘plateau’ when x → 0

Summary of Findings

The paper demonstrates that thermal activation generates a finite-size-controlled instability scale in disordered elastic media Dynamical heterogeneities reveal a twofold behavior where χ4 exhibits a maximum at a finite temperature T∗, separating intermittent dynamics from thermally assisted flow Avalanche statistics show that for T < Tc(L), power-law distributions with exponential cutoffs characterize intermittent dynamics, while for T > Tc(L), pronounced shoulders or peaks signal the emergence of runaway avalanches The critical temperature Tc depends on system size and decreases as L increases, suggesting that in the thermodynamic limit arbitrarily small but finite temperatures may destabilize the intermittent regime The results clarify that thermal activation alone can reorganize marginal stability and avalanche dynamics in amorphous solids

Acknowledgments

We would like to thank fruitful discussions with E.A. Jagla and A. Rosso and useful feedback from M. p. 1 Ozawa G.R.-L. would like to thank A. p. 1 B. Kolton, from Instituto Balseiro, for helpful discussions on the subject and for technical support with the CUDA implementation We acknowledge financial support from PIP 2021-2023 CONICET Project Nº 757 and the CNRS IRP Project “Statistical Physics of Materials” G.R.-L. acknowledges support from Universidad Nacional del Sur and FONCyT grant PICT-2021-I-A-01272 We acknowledge computational resources provided by the HPC cluster of the Physics Department at Centro At´omico Bariloche and the computing cluster at IFISUR, Bah´ıa Blanca

References

[1] P. Chauve, T. Giamarchi, and P. Le Doussal, Phys. Rev. B 62, 6241 (2000)

[2] A. B. Kolton, A. Rosso, T. Giamarchi, and W. Krauth, Phys. Rev. Lett. 97, 057001 (2006)

[3] A. Rosso, P. Le Doussal, and K. J. Wiese, Phys. Rev. B 80, 144204 (2009)

[4] E. E. Ferrero, S. p. 0 Bustingorry, A. B. Kolton, and A Rosso, Comptes Rendus Physique 14, 641 (2013)

[5] E. E. Ferrero, L Foini, T Giamarchi, A B Kolton, and A Rosso, Annual Review of Condensed Matter Physics 12, 111 (2021)

[6] J Lin, E Lerner, A Rosso, and M Wyart Proceedings of the National Academy of Sciences 111, 14382 (2014)

[7] C Liu, E E Ferrero, F Puosi, J L Barrat, and K Martens Phys. Rev. Lett. 116, 065501 (2016)

[8] A Nicolas, E E Ferrero, K Martens, and J L Barrat Rev. Mod. Phys. 90, 045006 (2018)

[9] E E Ferrero and E A Jagla Soft Matter 15, 9041 (2019)

[10] E E Ferrero and E A Jagla Phys. Rev. Lett. 123, 218002 (2019-11)

[13] M Ozawa and G Biroli Phys. Rev. Lett. 130, 138201 (2023)

[14] A Tahaei, G Biroli, M Ozawa, M Popovi´c, and M Wyart Phys. Rev. X 13, 031034 (2023)

[15] Y Takaha, H Mizuno, and A Ikeda Phys. Rev. E, (2025)

[16] T W J de Geus, A Rosso, and M Wyart Phys. Rev. E 111, L013503 (2025)

[17] D J Korchinski, D Shohat, Y Lahini, and M Wyart Phys. Rev. X 15, 031024 (2025)

[18] G Rodriguez-Lopez, K Martens, and E E Ferrero Phys. Rev. Mater. 7, 105603 (2023)

[19] A Cottrell, Journal of the Mechanics and Physics of Solids 1, 53 (1952)

[20] Here T∗(L) denotes the characteristic temperature (for a given system size L) at which we observe a crossover to a different dynamical regime

[21] A Nicolas, E E Ferrero, K Martens, and J L Barrat Rev. Mod. Phys. 90, 045006 (2018)

[22] G Picard, A Ajdari, F Lequeux, and L Bocquet The European Physical Journal E 15, 371 (2004)

[23] E E Ferrero, A B Kolton, and E A Jagla Physical Review Materials 5, 115602 (2021)

[24] M Popovi´c, T W J de Geus, W Ji, and M Wyart Phys. Rev. E 104, 025010 (2021)

[33] As a matter of fact, at low enough temperatures the time delays between avalanches of plastic activity are basically dominated by the thermal activation of the first site triggered

[34] The systematic choice of the weakest site for avalanche restarting is a good approximation at low temperatures but might eventually be replaced by a true probability tower of activations to have an accurate time clock (see e.g. [56])

[35] C Donati, S Franz, S C Glotzer, and G Parisi Journal of Non-Crystalline Solids 307-310, 215 (2002)

[36] L Berthier Physics 4, 42 (2011)

[37] K M Salerno, C E Maloney, and M O Robbins Phys. Rev. Lett. 109, 105703 (2012)

[38] K Karimi, E E Ferrero, and J L Barrat Phys. Rev. E 95, 013003 (2017)

[47] Z Budrikis and S Zapperi Phys. Rev. E 88, 062403 (2013)

[48] Z Budrikis, D F Castellanos, S Sandfeld, M Zaiser, and S Zapperi Nature Communications 8, 15928 (2017)

[49] J Lin, A Saade, E Lerner, A Rosso, and M Wyart Europhysics Letters 105, 26003 (2014)

[50] E E Ferrero and E A Jagla Journal of Physics: Condensed Matter 33, 124001 (2021)

[56] G Russo, E E Ferrero, A B Kolton, A Rosso, and D Vandembroucq Activation and avalanche length scales in the finite-temperature creep of an elastic interface (2026), arXiv:2604.

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. Thermal activation driven finite-size crossover in amorphous solids allows AI to predict system-spanning runaway events in disordered environments as temperature increases, moving beyond scale-free behavior. The improved AI can model complex material failure or phase transitions by predicting when a system will transition from intermittent dynamics to a fluidized state based on temperature and size.

  2. System size dependent critical temperature Tc(L) allows the AI to understand the stability limits of disordered systems, predicting that in the thermodynamic limit arbitrarily small but finite temperatures may destabilize the intermittent regime. This enables AI to identify conditions where purely thermal activation can generate an intrinsic instability scale in disordered elastic media without external driving.

  3. The analysis of dynamical heterogeneities through four-point susceptibility χ4 reveals correlated spatiotemporal rearrangements, which the AI can use to quantify local persistence measurements and reveal the emergence of correlated spatiotemporal rearrangements as temperature is varied. This capability allows the AI to better characterize how local activation rules lead to system-wide reorganization.

  4. The understanding of avalanche statistics shows that for T < Tc(L), avalanches have power-law distributions with exponential cutoffs, while for T > Tc(L), they develop pronounced shoulders or peaks, signaling the emergence of runaway avalanches and the onset of thermally assisted flow. This enables the AI to distinguish between scale-free intermittent events and system-spanning runaway avalanches in complex simulations.

  5. The relationship between avalanche size and duration resembles that in sheared systems, which allows the AI to use this correspondence to predict avalanche characteristics even when purely thermal activation is present, providing insights into thermal activation alone can generate a finite-size-controlled instability scale.

Sources

Related papers