Precursory information for rock failure based on the principle of critical slowing down

Abstract

In this paper, laboratory experiments of rock under tension and compression were carried out, including the uniaxial compression test and Brazilian test. Real-time acoustic emission (AE) monitoring was conducted during the deformation and failure process of rock. By introducing the critical slowing down theory, we analyzed AE count sequences during the whole loading process. The results show there exists critical slowing down feature of AE signals prior to rock failure. There is a sudden and significant growth in the variance and autocorrelation coefficient of acoustic emission preceding rock failure, which can be viewed as precursory indicator for rock failure. The precursory indicators acquired from the variance are more obvious than the autocorrelation coefficient. The time intervals between the precursory indicator and rock failure moment are less and the ratios of the precursory indicator to rock failure moment are higher under tension compared to compression. The findings can improve the development of the acoustic emission monitoring technology.