Modeling transient excavation-induced dynamic responses in rock mass using an elasto-plastic cellular automaton

The problem of transient excavation-induced dynamic response in rock mass is discretized on both spatial and temporal scales and is solved by using the cellular automaton (CA) technique and the Newmark scheme, respectively. Using this approach, a dynamic analysis version of 3D elasto-plastic cellular automaton (EPCA(3D)) is developed. The advantage of this method is it avoids the solution of large-scale linear equations since only the local CA rule is used for dynamic state updating. The Den Iseger algorithm is introduced to validate the numerical method. The evolution of stress and velocity for the transient excavation in rock mass under different conditions obtained by the EPCA(3D) and Den Iseger methods are in good agreement. The abilities of EPCA(3D) in the modeling of elasto-plastic dynamics of transient excavation in rock mass are well demonstrated. By considering different in situ stresses, excavation radiuses, unloading durations and unloading paths, the factors affecting nonlinear dynamic responses are investigated. It is found that the failure extent increases with the decrease of unloading time. When lateral pressure coefficient (sigma(x)/sigma(y)) equals to 1, the failure zone is evenly distributed around the tunnel. However, with the decrease of lateral pressure coefficient, the extent of failure localization increases. The fault around the tunnel makes the dynamic failure zone asymmetric distribution. The modeling helps to understand the major mechanism of rock mass damage for the transient excavation.