Identification and early warning method of key disaster-causing factors of AE signals for red sandstone by principal component analysis method

Abstract

The failure process of rocks is usually accompanied by numerous acoustic emission (AE) signals. For evaluating the damage state of rocks, it is important to select key disaster-causing AE signals from massive AE monitoring data. Based on the uniaxial compression test of red sandstone, the failure characteristics of AE signals were analyzed. Then, based on the principal component analysis method, the evolution pattern of key disaster-causing factors of AE signals of red sandstone has been obtained. The results demonstrated that AE parameter signals (AE energy and AE ring counts) and AE waveform signals (AE main frequency) contributed to characterize the precursor of rock failure. The values of AE energy and AE ring counts increased significantly in the critical failure stage of rocks, and there existed a short quiet period phenomenon of AE signals. Similarly, AE main frequency increased densely in the critical failure stage. Moreover, the optimized key disaster-causing signals (AE energy, AE ring counts and AE main frequency) can clearly characterize the stress fluctuation and damage state of rocks. Based on the optimized key disaster-causing AE signals, the primary early warning point and the key early warning point of rockburst were proposed. We hope this method can bring some new ideas for predicting rockburst.

Keywords

• AE signals
• Disaster-causing factors
• Early warning method
• Rockburst