AUTHOR=Zhu Xing , Wang Luqi , Yang Yang , Zhang Wengang , Zhang Peng 

TITLE=Cracking Modes and AE Precursors of Sandstone Failure Under Multi-Stage Uniaxial Compression

JOURNAL=Frontiers in Earth Science

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2022.874543

DOI=10.3389/feart.2022.874543

ISSN=2296-6463

ABSTRACT=The micro-cracking mode (tension and shear) of rock is constrained by the internal physical mechanism, identification of microcrack modes and their evolution characteristics provide a promising tool for determining the destructive state of a rock sample and can provide early warning of final failure. In this paper, a multi-stage uniaxial compression loading test was carried out on sandstone rock samples. The AE (Acoustic Emission) signals released by cracking sandstone samples under the entire loading process are collected in waveform and parameters. Then, an unsupervised machine learning approach k-means classification is applied to classify all AE event signals into two clusters corresponding to tensile and shear cracking modes from AE parameters similarity. The results demonstrated that the AE signals produced by the cracking of sandstone samples have obvious stage characteristics, and the k-means method is a viable option to differentiate the cracking models, and rise time was the key parameter in distinguishing the cracking modes. Furthermore, the autocorrelation coefficient and the variance of the rise angle (RA) of AE events were studied as the critical transition indicators of rock destructive states based on the critical slowing down (CSD) theory. It was found that there is an obvious CSD phenomenon approaching final rock failure, which is manifested by a significant increase of the autocorrelation coefficient and the variance of RA. It is worth noting that the start time of CSD is almost the same as the time of transition of mode from tensile to shear in this study. Compared with the autocorrelation coefficient, the variance of RA is a better indicator of rock destruction and provides obvious precursory information of rock damage and failure.