Xu, Y-H, Esterhuizen, G, Jakubec, J, Russo, A & Thomas, A 2026, 'Practical fragmentation estimates using a new block cave fragmentation tool', in A van As, D Cumming-Potvin & J Wesseloo (eds), Caving 2026: Proceedings of the Sixth International  Conference on Block and Sublevel Caving, Australian Centre for Geomechanics, Perth, pp. 1-15, https://doi.org/10.36487/ACG_repo/2645_32
(https://papers.acg.uwa.edu.au/p/2645_32_Xu/)
Abstract: Realistic fragmentation estimates are a critical input for block cave mine design, yet they remain a great challenge for mining practitioners. The block cave fragmentation (BCF) software has been utilised in cave mining design since the 1990s. Although realistic input parameters are key to obtain realistic results in any methodology, fragmentation predictions using BCF or any available tools are challenging especially in predicting the smaller rock fragments and fines (<5 cm). This is largely due to a scarcity of field data from operating caves required to calibrate prediction logic. Consequently, the proportion of fines is often underestimated, despite the introduction of sophisticated approaches such as explicit modelling using discrete fracture networks (DFN) and numerical codes. The authors discuss the reasons for the poor correlation between predictions and operational reality across all 3 fragmentation stages: in situ, primary, and secondary. This paper reviews initial experiences with the newly released BCF version 4 (V4), which attempts to address these shortcomings by improved simulation of the crushing/grinding/block splitting processes in a draw column. Significantly, the updated BCF has been calibrated against fragmentation data collected from mines with diverse rock mass characteristics at varying depths, including jointed, weak, and veined ground. The paper also outlines a refined workflow, tested against real-world cases, that incorporates DFN and threedimensional fracture-mechanics-based hybrid codes (3D-FDEM).

Keywords: block caving, fragmentation, fines predication, DFN, 3D-FDEM