Document Type : Research Paper
Authors
1
Department of Geology, Faculty of Science, Ferdowsi University of Mashhad, Mashhad, P.O. Box 91775-1436, Iran
2
Professor, Department of Geology, Faculty of Science, Ferdowsi University of Mashhad, Mashhad, P.O. Box 91775-1436, Iran.
3
Professor, Department of Geology, Faculty of Science, Ferdowsi University of Mashhad, Mashhad, P.O. Box 91775-1436, Iran
4
Assistant Professor, Department of Mining Engineering, Faculty of Engineering, Tarbiat Modares University, Tehran, Iran.
5
Assistant Professor, Civil Engineering Department of Civil Engineering, Shahine., Islamic Azad University, Shahinshahr, Iran.
Abstract
Abstract:
Accurate prediction of rock drillability is critical for optimizing tunneling, mining, and excavation operations in hard rock environments. This study investigates the drillability of granitic rocks by integrating petrographic, physical, mechanical, and abrasivity parameters with both conventional regression and machine learning (ML) approaches. Laboratory tests were conducted on samples from six granitic rock groups, measuring properties such as brittleness, hardness, abrasivity indices, and the Drilling Rate Index (DRI). Statistical analyses, including linear and nonlinear regression, revealed strong nonlinear relationships between DRI and engineering parameters, with R² values up to 0.95. Machine learning models, particularly Random Forest (RF) and Artificial Neural Networks (ANN), were applied independently, with RF achieving superior predictive performance (R² > 0.99) and lower error indices compared to ANN and regression models. The study also highlights the influence of different rock groups on model performance and discusses limitations related to dataset scale, in situ conditions, and model generalization. These results confirm that integrating laboratory measurements with ML techniques provides a reliable framework for predicting rock drillability, offering practical guidance for excavation planning, equipment selection, and operational efficiency.
Keywords: Granitic rocks, Drilling Rate Index (DRI), Machine Learning, Random Forest, Artificial Neural Networks, Rock drillability, Abrasivity, and Rock mechanics
Keywords
Main Subjects
)