Critical slowing down precursor information for the acoustic emission response characteristics of defective tuffs

Fracture characteristics of rock and rational structural design are necessary to ensure the stability of underground rock engineering. Early warning of rock damage is a hot and challenging issue for research in geotechnical engineering. In this paper, uniaxial compression tests and acoustic emission (AE) tests were conducted on different defective specimens. The AE characteristics during rock damage were statistically analyzed based on the critical slowing theory. The results showed that the stress fluctuation in the stress-strain curve appeared earlier with the increase of initial damage. The failure pattern shifted overall from tensile to shear cracks. With the load increased, the AE amplitude changed from low amplitude and low density to high amplitude and high-density characteristics. The AE RA/AF was more sensitive to tiny cracks and had better immunity to attenuated AE signals. Both cumulative AE energy and RA/AF showed a precursor characteristic of a sudden increase in slope before rock damage. The AE b-value experienced a period of rising, fluctuation, and decline, with a sudden drop before the large-scale crack extension, showing a decreasing trend overall. The magnitude of fluctuations and density of distribution of the critical slowing down parameters (variance and autocorrelation coefficients) re-flected the severity of specimen failure. The early warning characteristic of the variance was more significant, while the autocorrelation coefficient had an apparent disorder, which interfered with the judgment of the system phase transition. The average warning time of different AE parameters' "critical precursor point" was greater than 42 s, and the stress level was lower than 91.50%. The warning time sequence of different parameters in descending order was as follows: variance of AE parameters > autocorrelation coefficient of AE parameters > traditional AE parameters. The overall warning effect of the critical slowing down phenomena was better than that of classic AE parameters.