Simulation of the fragmentation and propagation of jointed rock masses in rockslides: DEM modeling and physical experimental verification

Rockslides are among the most dangerous geohazards. The discrete element method (DEM) has been increasingly used to simulate the emplacement process of rockslides. However, few studies have modeled the fragmentation of jointed rock masses during movement via DEM simulations and verified the reliability of the results through comparison with the results of laboratory experiments and field events. In this paper, a series of DEM simulations based on PFC3D5.0 was conducted to replicate laboratory experiments in which breakable blocks with different structural forms impact and slide on a horizontal plane and experience fragmentation. The results show that the DEM model provides first-order estimates that are comparable with the results of experiments in terms of the kinetic features and depositional characteristics of rockslides, including crack propagation, velocity distribution, preservation of source stratigraphy, progressive fragmentation, and spreading of fragments. Conversely, the normalized horizontal runout and degree of fragmentation are poorly represented in the DEM model. A theoretical analysis is further conducted to discuss the causes of travel distance discrepancy between the DEM model and experiments. It is found that the inconsistency of fragment interactions caused by the defect of the fragmentation process in DEM simulation may take primary responsibility. An alterable restitution coefficient method based on the rock mass fragmentation process is proposed for application to the fragmenting rock mass movement in DEM simulation, with the utilization of additional laboratory experimentation as the calibration benchmark. Further verification of our method needs to be performed in the future.