Improvements in blast fragmentation models using digital image processing

This paper presents an approach for developing accurate models to predict the fragmentation due to blasting. The approach makes use of drill monitoring data, which provides data throughout the rock mass to be blasted and is therefore superior to traditional data-gathering methods such as diamond core or field sampling. The approach also makes use of the Split image processing software for assessing post-blast fragmentation and the crushability and grindability of the ore. Finally the approach makes use of the explosive energy per unit volume of rock. These three types of data are collected and analyzed on a hole-by-hole basis, giving 50 or more data points for each blast. These data points form the basis for a statistical correlation between in-situ conditions, blasting parameters, and the resulting fragmentation size and strength. At a specific mine, the database is continually updated as mining progresses, resulting in an evolving and increasingly accurate model with time. Some sample results from the Phelps Dodge Sierrita mine are presented.