Prediction of Laser Penetration Rate in Volcanic Rocks Based on Physicomechanical and Mineralogical Properties
Journal of Materials Engineering and Performance, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Basım Tarihi: 2026
- Doi Numarası: 10.1007/s11665-026-15298-5
- Dergi Adı: Journal of Materials Engineering and Performance
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Applied Science & Technology Source, Chemical Abstracts Core, Compendex, INSPEC, Engineering Source (EBSCO), Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest)
- Anahtar Kelimeler: clustering, laser penetration rate, P-wave velocity, regression, volcanic rocks
- Ankara Üniversitesi Adresli: Evet
Özet
Laser drilling has gained increasing attention in hydrocarbon exploration, geothermal energy, and mining applications. However, predicting laser penetration rate (LPR) directly from the intrinsic properties of heterogeneous volcanic rocks remains challenging under low-power laser conditions, as LPR is governed by the combined effects of physicomechanical, mineralogical, geochemical, and textural properties. In this study, global regression models applied to the entire dataset (n = 24) explained only limited variance (R2 ≤ 0.35), highlighting the complex and heterogeneous controls on LPR. To overcome this, a clustering-assisted multivariate regression framework was developed that first partitions the dataset into statistically homogeneous subgroups prior to regression, an approach not previously applied to laser–rock interaction. Within the resulting 22-sample subgroup, P-wave velocity emerged as the strongest single predictor (R2 = 0.89), while a five-variable exponential model achieved R2 = 0.97. Colorimetric (RGB) features alone proved insufficient (R2 ≤ 0.30), confirming that optical screening cannot substitute for integrated multi-parameter characterization. Across 631 evaluated regression and clustering models, this framework consistently outperformed conventional global regression. These findings demonstrate that clustering-assisted, multi-parameter modeling substantially improves LPR predictability and provides a practical, property-based tool for preliminary machinability assessment, reducing reliance on extensive laser trial testing in rock excavation planning.