ABSTRACT
Rock burst is a progressive failure process of microcrack initiation, propagation and coalescence. Routine monitoring methods are unable to predict rock bursts since the evolution of microcracking is difficult to record in hard rock mass. In this paper, the feasibility of predicting rock bursts is discussed by considering source zones, development cycle and scale. The authors think the feasibility can be understood in three aspects: (1) the heterogeneity of rock is the main reason for the existence of rock burst precursors; (2) deformation localization is the intrinsic cause of rock bursts; and (3) the interaction between the target rock mass and the surrounding rock mass is the external cause of rock bursts. As an example, the microseismic monitoring used in the tunnel construction of Jinping II Hydropower Station is reported. It is found that the rock burst development is a static process and the rock burst occurrence is within the scope of dynamics. In addition, some precursors exist before rock bursts occur. This is reflected by the consistency between the acoustic emissions and the damages in the rock mass. There are three seismogenic models according to the tempo-spatial distribution. By utilizing the 3S (Stress buildup, Stress shadow and Stress transfer) principle in seismology, three rock burst criteria are proposed in this paper (namely, the stress tempo-spatial evolution, the magnitude, energy and concentration degree of microseismic events, and the sudden change in the apparent volume). The effectiveness of using the microseismic method to predict rock bursts has been validated in some typical engineering cases. The results of this study may be valuable for improving the accuracy of rock burst prediction, and be helpful in minimizing equipment damage and injury of personnel during construction of deep-buried tunnels.
