The dynamic behavior of crack propagation and coalescence in rock is investigated using the distinct lattice spring model (DLSM). Discontinuum based element model has been proven as a promising tool in the study of dynamic fracture in brittle materials like rock. As a newly developed tool, the distinct lattice spring model (DLSM), compared with classical discrete element model (DEM), has advantages on computationally efficiency and being free of micro parameters calibration. This study further applies the DLSM to investigate the dynamic cracking and coalescence behavior in brittle material. First, central notched Brazilian discs (CNBD) were simulated under loading. The results on crack pattern and coalescence are in good agreement with the experimental observations. The 3D examination of the evolution of the crack showed the crack growth in 3D was not as 2D observation, e.g. crack propagation front is very irregular and failure might appear ahead of the crack tip. Second, the simulations on a perforated plate blasting loading reveals that the crack pattern highly relays on the mode of heterogeneity. More specifically, the micro-structure of the material plays an important role on dynamic crack propagation and coalescence. At last, blasting effect on a 3D volume has also been investigated. This study can be useful to provide better understanding of the failure of rock under dynamic loading with discontinuity or heterogeneity.
Crack propagation and coalescence in rock is a popular research topic especially in underground excavation projects or slope stability problems in which the dynamic fracture phenomena are in dominance. To reduce the scale of the problem, different models are designed for experimental studies (e.g. perforated thin plate model  and pre-notched disc model ). These models focus on the most important feature of the crack during fracturing by setting up specific specimen configuration or loading condition. With the development of the modern technology, numerical methods have been improved a lot. Following the experimental model, numerical methods are applied on these models for validation or fundamental studies prior to applications on complicated project simulation.