Mesoscopic fluctuations and intermittency in aging dynamics

TL;DR

This paper presents a model linking intermittent rearrangements to aging dynamics, showing how autocorrelation decays and its distribution evolves, validated against spin glass simulations.

Contribution

It introduces a simple model connecting intermittent events to aging autocorrelation decay and distribution shape, supported by simulation data.

Findings

Autocorrelation decays algebraically with aging.

Probability density function approaches Gaussian at large times.

Model aligns well with Edwards-Anderson spin glass simulations.

Abstract

The configurational de-correlation in an aging system is attributed to irreversible intermittent rearrangements, which are described as a Poisson process with average $\propto \ln(1 + t/t_w)$, where $t$ is the observation time and $t_w$ is the age [P. Sibani and H.J. Jensen, Europhys. Lett. 69, 2005]. On this basis, we obtain a simple model for the off-equilibrium aging behavior of the autocorrelation: the average autocorrelation decays algebraically, and its shifted and rescaled probability density function (PDF) has a Gumbel-like shape which approaches a Gaussian at large times and becomes sharp in the thermodynamic limit. The model properties are tested against simulations of the Edwards-Anderson spin glass, and are in reasonable agreement with other available data.