Machine Learning-Based Estimation of Unconfined Compressive Strength and Deformation Modulus of Heated Gold Coast Sandstones

DOI

10.2991/978-94-6463-900-1_37How to use a DOI?

Keywords

Machine learning; Sandstones; Unconfined compressive strength; Elastic modulus; Soft computing

Abstract

Unconfined compressive strength (UCS) and elastic modulus (E) are crucial in underground structures such as deep mining, tunnels, geothermal energy systems, and nuclear waste disposal repositories. It is inevitable to encounter a rock mass with weak and heavy geological discontinuities, which makes it challenging to obtain high-quality rock cores to determine UCS and E. Furthermore, determining UCS and E through laboratory tests is costly and laborious; hence, it is valuable to accurately estimate UCS and E of Gold Coast sandstone using inexpensive and more reliable techniques. This study employed 3 machine learning models: support vector machine (SVM), random forest (RF), and extreme gradient boost (XGBoost) proposed for rock strength prediction. A total of 87 laboratory data sets from 2 sandstones were used to develop the proposed models. The models consider porosity (η), temperature (T), UCS, E, and density (ρ) as model input parameters. The results demonstrated that the accuracy of the SVM model surpasses the other two models, with a correlation coefficient R2 of 0.99 for UCS and E each. The Shapley additive explanations (SHAP) and the correlation matrix analysis indicate that dry density and porosity are the primary input factors, with temperature being the least influential factor in predicting UCS and E. The available data shows that predictive models are proven reliable in estimating UCS and E of the Gold Coast sandstones.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Stephen Akosah
AU  - Ivan Gratchev
AU  - Qianhao Tang
PY  - 2025
DA  - 2025/12/07
TI  - Machine Learning-Based Estimation of Unconfined Compressive Strength and Deformation Modulus of Heated Gold Coast Sandstones
BT  - Proceedings of the Rocscience International Conference 2025 (RIC 2025)
PB  - Atlantis Press
SP  - 377
EP  - 386
SN  - 2589-4943
UR  - https://doi.org/10.2991/978-94-6463-900-1_37
DO  - 10.2991/978-94-6463-900-1_37
ID  - Akosah2025
ER  -