Coupled geomechanical classification and multivariate statistical analysis approach for the optimization of blasting rock boulders

Abstract

The prediction of blasting rock boulder in discontinuous rock is crucial for the optimization of blasting operations. Therefore, it is necessary to understand the role played by the geological features of a rock mass in controlling oversize fragments. This research is the result of coupled use of multivariate analysis methods and geomechanical indexes in order to identify the main rock mass and blast parameters that affect directly the oversize boulder production during blasting operations. The aggregate quarries selected for this study belong to the Eocene and the Jurassic rock in Tunisia. Diverse techniques were used as atomic absorption, XR diffraction, microscopic study, mechanical test, and scanning electron microscopy image to identify rock matrix. The methodology established by cluster analysis generated from mechanical classification as rock quality designation, rock mass rating, Q-Barton index, and strength index makes possible to classify the studied rock into three classes. A principal component analysis method developed in XLSAT 2018 has been performed on various blast design parameters to illustrate the relation between blasting and rock parameter. For the prediction of oversize fragment resulting from the blasts, a specific formula for every quarry was generated by statistical method. The proposed formulas can be considered as sufficient with an accuracy of more than 80% of the blasted rock after model testing compared with precise boulder percent in twenty blasts.