Abstract
Predicting and optimising blast-induced rock fragmentation is essential for improving downstream mining operations such as loading, hauling, and comminution. The widely used Kuz-Ram model often requires heuristic calibration of the rock factor, which limits its predictive accuracy in heterogeneous rock masses. This study introduces a modified Kuz-Ram model that integrates response surface methodology (RSM) to refine the rock factor estimation based on the blastability index (BI). A dataset of 80 production blasts from Orapa Diamond Mine, Botswana, incorporating four input parameters, powder factor, charge, rock factor, and blastability index was analysed. The RSM-modified model demonstrated a 12.2% improvement in prediction accuracy compared to the traditional Kuz-Ram model, achieving an R2 of 0.579, with RMSE and MAE reduced to 5.71 and 3.07 respectively. The approach preserves the empirical simplicity of Kuz-Ram while explicitly modelling nonlinear parameter interactions, offering a transparent yet robust tool for blast design. This method is practical for mines requiring rapid calibration to site-specific geological conditions and can be extended to include additional parameters, such as delay timing, in future work.
| Original language | English |
|---|---|
| Pages (from-to) | 105-112 |
| Number of pages | 8 |
| Journal | International Journal of Mining and Geo-Engineering |
| Volume | 60 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 2026 |
| Externally published | Yes |
All Science Journal Classification (ASJC) codes
- Geotechnical Engineering and Engineering Geology
- Geology
- Economic Geology
Fingerprint
Dive into the research topics of 'Modification of the Kuz-Ram model using response surface methodology to optimise blast fragmentation'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver