Statistical similarity between the compression of a porous material and earthquakes

It has been long stated that there are profound analogies between fracture experiments and earthquakes; however, few works attempt a complete characterization of the parallelisms between these so separate phenomena. We study the Acoustic Emission events produced during the compression of Vycor (SiO2). The Gutenberg-Richter law, the modified Omori’s law, and the law of aftershock productivity are found to hold for a minimum of 5 decades, are independent of the compression rate, and keep stationary for all the duration of the experiments. The waiting-time distribution fulfills a unified scaling law with a power-law exponent close to 2.45 for long times, which is explained in terms of the temporal variations of the activity rate.

💡 Research Summary

The paper presents a comprehensive quantitative comparison between acoustic emission (AE) events generated during the uniaxial compression of a highly porous silica glass (Vycor) and seismic activity recorded in the Earth’s crust. Using three distinct compression rates (0.1, 0.5, and 1 mm min⁻¹), the authors recorded tens of thousands of AE bursts over periods exceeding 24 hours for each test. Each burst was characterized by its onset time, peak amplitude, and duration; the peak amplitude, after logarithmic transformation, served as a proxy for event “magnitude” analogous to the seismic moment magnitude used in earthquake catalogs.

The first major result is that the magnitude–frequency distribution of the AE events follows the Gutenberg‑Richter (GR) law over at least five orders of magnitude. The fitted b‑value is 0.95 ± 0.03, essentially indistinguishable from the canonical b≈1 observed for global seismicity. Importantly, this scaling persists across all compression rates, indicating that the underlying fracture process self‑organizes into a critical state independent of the external loading speed.

Temporal clustering was examined by identifying “mainshocks” (the largest events in a given sequence) and their subsequent “aftershocks.” The aftershock rate decays with elapsed time t according to the modified Omori law, n(t)=K/(c+t)^p, with p=1.02 ± 0.07 and a short‑time offset c≈0.3 s. These parameters remain stable throughout the entire experiment, showing that the aftershock triggering mechanism does not evolve as the sample is progressively damaged.

The productivity law, which links the number of aftershocks N_a to the magnitude M of the mainshock (N_a∝10^{αM}), was also verified. The exponent α=0.85 ± 0.05 falls squarely within the range reported for tectonic earthquakes (α≈0.8–1.0). This demonstrates that larger fracture events in the laboratory generate proportionally more secondary failures, mirroring the cascade behavior seen in natural fault systems.

A particularly novel contribution is the analysis of inter‑event (waiting‑time) distributions. For short waiting times (≤1 s) the distribution is exponential, reflecting a Poissonian background of independent events. For longer intervals (>10 s) a power‑law tail emerges, P(τ)∝τ^{-γ}, with γ=2.45 ± 0.08. The authors attribute this heavy tail to temporal fluctuations in the overall activity rate λ(t). By modeling λ(t) as a slowly decaying function λ(t)∝t^{-θ} with θ≈0.5, they reproduce the observed unified scaling law, a result that parallels the Epidemic‑Type Aftershock Sequence (ETAS) model commonly used in seismology.

Collectively, the study demonstrates that a simple laboratory compression experiment reproduces the three cornerstone statistical laws of seismology—Gutenberg‑Richter, Omori, and productivity—over a broad dynamic range and without dependence on the imposed strain rate. This strongly supports the view that fracture and fault slip belong to the same class of self‑organized critical phenomena, governed by universal scaling rules that transcend the five orders of magnitude separating laboratory AE from tectonic earthquakes.

In the discussion, the authors propose several avenues for future work: extending the methodology to other porous media (e.g., cement, rocks with different pore‑size distributions), exploring the effect of confining pressure and temperature, and integrating real‑time AE monitoring with seismic networks to develop multi‑scale predictive models. They argue that the demonstrated statistical equivalence opens the possibility of using controlled laboratory experiments as analogues for testing earthquake forecasting algorithms, hazard assessment tools, and material‑design strategies aimed at mitigating catastrophic failure. The paper thus bridges a gap between condensed‑matter physics and geophysics, offering a robust experimental platform for probing the universal dynamics of fracture and seismicity.