Critical slowing down features of acoustic emission signals for predicting the failure of black sandstone under different freezing temperatures

This study examines the influence of sub-zero temperatures on the mechanical behavior and failure prediction of black sandstone. For this, quasi-static compression tests were conducted on black sandstone specimens under various temperatures, 5 °C, -5 °C, -10 °C, and − 20 °C. An acoustic emission (AE) monitoring technique was utilized to reveal the damage features of the rock at negative temperatures. The autocorrelation coefficient (AC) and variance of AE counts were assessed using the critical slowing down (CSD) theory to examine the precursor characteristics of rock failure under sub-zero temperatures. Further, the analysis of correlation dimension (CD) evolution was conducted to validate the results of CSD theory. The results indicate that as the temperature decreased from 5 °C to -20 °C, the uniaxial compressive strength (UCS) of the black sandstone increased by 43.09%. The AE counts, and cumulative counts effectively reflect the damage progression in the rock under compressive loading. The AE counts, and AE cumulative counts gradually rise with decreasing temperatures, indicating a more intense AE response. The AE signals associated with rock failure demonstrate CSD phenomena, where abrupt increases in the AC and variance curves of AE counts can be used to predict the ferocious failure. Furthermore, the findings show that the precursory time lag in black sandstone samples increases as the temperature decreases. Compared to CD and AC curves, the variance curve of AE counts provides a more distinct early warning feature for predicting rock failure under sub-zero temperatures. Consequently, this research holds significant implications for the prediction of rock failure in cold regions.