New computational framework for modeling the gravity flow behavior of sublevel caving material

Abstract A better understanding of the gravity flow mechanism of caved material is required to increase ore recovery in caving mining. This study presents a novel computational framework PRF3D (3D particle random flow) capable of simulating the flow behavior of caved ore and rock in sublevel caving mining. The PRF3D framework combines the advantages of the discrete element method and stochastic medium theory and uses the billboard technique of OpenGL to render and visualize particles. The new framework can automatically record the ore recovery indicators, display the shape of the draw zone in real time and capture the fines percolation. An isolated physical draw experiment is conducted to validate the accuracy and efficiency of the PRF3D from both qualitative and quantitative perspectives. After validation, various numerical tests are conducted to investigate the influence of the sublevel height, drawpoint spacing, drawpoint size and particle size of overlaying rock on the multidraw behavior of caved ore and rock in sublevel caving mining. The performance of PRF3D is also demonstrated by benchmarking with the reported simulations.