Abstract: In order to accurately grasp the flow characteristics of ore and rock of block caving， a new method that combining attribute block model with stochastic medium theory was proposed. The procedure of simulation flow is as follows: attribute block model was established to fit and discretize target region. Block attributes were obtained by attribute interpolation technique. A proper flow model for programming was established. The concept of draw-out index and block index was put forward to analyze the flow characteristics of the rock under different oredraw conditions. Flow simulation was realized based on the proposed method in 3D visualization platform with C++ language. The results show that with the draw height increasing， the length of the long axis of the draw out body increases linearly， and the length of the short axis and the eccentricity increase gradually， showing a power exponential relationship. Under constant block index， with the increase of draw-out index， the ellipsoid of the ore body is gradually developed， the eccentricity ratio is increased， and the average depth of the depression increases. Under constant draw-out index， with the increase of block index， the eccentricity of the ellipsoid is decreased.

Key words: attribute block model, ore and rock flow model, stochastic medium, particle flow, block caving