Abstract
We investigate the mechanisms for rock-burst using the numerical Discontinuous Deformation Analysis (DDA) method. Using recently developed non reflective boundary and excavation sequence modeling capabilities we are now able to model dynamic deformation in high in-situ stress environments more accurately than before. First we perform verifications of P-wave propagation through a one-dimensional elastic bar and confirm DDA accuracy provided that the block length with respect to wave length is properly conditioned. We then test a newly developed radial P-wave propagation module to emulate an underground blast. We study two possible rock burst generation mechanisms: 1) due to strain relaxation as response to opening in high in situ stress environment, and 2) due to nearby blasting. A very strong relation between the initial stress and the velocity and acceleration of the ejected key blocks following the removal of the tunnel section is reported. We also find that the influence of blasting on rock burst phenomena is strongly related to the initial in situ stress level. We conclude that under relatively low in situ stress environments nearby blasting may indeed ejection of originally stable key blocks. However, under high in situ stress conditions strain relaxation poses a much greater rock-burst risk.
1. INTRODUCTION
A rock-burst is the sudden displacement of rock block that occurs in the boundary of an underground excavation, and causes substantial damage to the excavation [1]. Due to the great risk to worker safety and the extensive damage to equipment, rock-bursts are considered to be one of the biggest unresolved problems in deep excavations.
There are two basic rock-burst mechanisms: 1) Strain relaxation that leads to the displacement of excavation surfaces, in which case the source and damage are concurrent, and 2) Seismic wave propagation from energy redistribution that is induced by explosions and drilling in the excavation, in which case the source and the damage might be separate in distance and time [2].
