Investigation of rock blast fragmentation based on specific explosives

In order to control and optimize a mining operation, it is critical to assess fragmentation caused by blasting and subsequent crushing and grinding stages. Blasting engineers are challenged to predict the mean size of a fragmented rock based on the characteristics of the rock mass, blasting geometry, technical parameters and explosive properties. Some of the effective parameters for rock fragmentation have been investigated in several empirical models. A model for fragmentation in bench blasting was developed using the effective parameters in the existing empirical models. It proposes a simple model for predicting the X50 value. The proposed model was calibrated by nonlinear fits to 35 bench blasts at different sites at the Sungun copper mine, the Akdaglar quarry, and the Mrica quarry. The developed model was validated by comparing it to data obtained from six blast sites at the Chadormalu iron ore mine and the Porgera gold mine. The results indicated a small variance in X50, which was calculated by the proposed model through an image processing approach. In comparing the proposed and Kuz-Ram models, the specific explosive energy and powder factor are almost identical. The advantage of the proposed model over the Kuz-Ram model is the specific explosive energy, since this parameter includes the powder factor and the weight strength of an explosive. In addition, a sensitivity analysis was conducted based on an artificial neural network. The results showed that the burden and the specific explosive energy were the most effective parameters in the designed model.