Non-contact laser vibrometry for rock failure precursor analysis: a comparative study with AE methods

To explore the precursory characteristics of failure in fractured sandstone. This paper uses laser vibration measurement technology and acoustic emission (AE) technology to monitor the failure process of red sandstone. The wavelet packet energy entropy (WPEE) method was used to process rock vibration data, indicating a sharp increase in WPEE as a precursor to rock failure. Concurrently, ring count, b-value, variance, and autocorrelation coefficient (AC) methods were applied to analyze AE precursors.The results show that ring count, variance, and AC increase significantly while the b-value decreases sharply prior to rock failure. These changes can all be used as precursor information of rock failure. Subsequently, in order to compare the advantages of different methods, we proposed a comparative framework that includes the timeliness, stability, and discriminability of predictions. The results indicate that the discriminability and stability of ring count and variance are the best, but their timeliness is the worst. The timeliness of b-value and autocorrelation coefficient is the best, but their stability and discriminability are the worst. The timeliness, stability, and distinguishability of WPEE are all in the middle position, and its overall performance is relatively good compared to other methods.