Superstatistical approach of electric potential and acoustic emission for investigating damage evolution and precursor of water-bearing sandstone under uniaxial compression

Abstract

Deep resource mining and geotechnical structures are prone to serious geological disasters in complex environments of abundant groundwater and high geostress. Comprehensive analysis of electric potential (EP) and acoustic emission (AE) from a superstatistical perspective promises predictive insights, necessitating further exploration. In this paper, to reveal the effects of different water contents and acidity, the EP and AE monitoring tests of coarse sandstone under uniaxial compression were conducted. The temporal evolution and superstatistical features of AEs and EPs were analyzed to elucidate correlations with pore-fissure parameters and mechanical properties. The impact of water content and acidity on response mechanisms of EPs and AEs was investigated by nuclear magnetic resonance (NMR), scanning electron microscope (SEM) and X-ray Diffraction (XRD) tests. Based on Tsallis entropy q_EP and q_AE, a novel muti-geophysical prediction method were proposed. The results indicate mechanical properties and AE activities decrease with increasing water content and acidity, while the average EP value enhances due to electrokinetic effect. EPs and AEs exhibit superstatistical features, and q_EP and q_AE decrease with increasing water content and acidity, correlating negatively with porosity and pore fissure dimensions. It is attributed to reduced frictional slip and rupture severity by water lubrication and chemical erosion, diminishing charge generation and elastic wave release. q_EP and q_AE vary from 1 to 3, and display drastic fluctuation after plastic stage. Then damage weighting coefficient and damage correlation coefficient proposed are used to identify the precursory of rock failure and obtain good prediction effect. These findings offer new perspectives for the prediction of geological disasters associated with water-rich environments.