Assessment and Prediction of Rock Drillability in Hard Granitic Rocks Using Experimental Testing and Machine Learning Models

Articles in Press

Document Type : Research Paper

Authors

1 Department of Geology, Faculty of Science, Ferdowsi University of Mashhad, Mashhad, P.O. Box 91775-1436, Iran

2 Department of Mining Engineering, Faculty of Engineering, Tarbiat Modares University, Tehran, Iran

3 Department of Civil Engineering, Shahi.C., Islamic Azad University, Shahinshahr, Iran

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

Main Subjects

Article Title [Persian]

_

Geopersia

Articles in Press, Accepted Manuscript
Available Online from 24 January 2026

Files

History

  • Receive Date: 01 September 2025
  • Revise Date: 15 January 2026
  • Accept Date: 24 January 2026

Share

How to cite

Statistics