Artificial neural networks to predict deformation modulus of rock masses considering overburden stress. Issue 1 (2nd January 2023)

• Record Type: Journal Article
• Authors: Tokgozoglu, K.; Aladag, C. H.; Gokceoglu, C.
• Is Part Of: Geomechanics and geoengineering. Volume 18:Issue 1(2023)
• Page Start: 48
• Page End: 64
• Publication Date: 2023-01-02
• Subjects: Deformation modulus -- rock mass -- intact rock -- ANN -- overburden stress; Engineering geology -- Periodicals; Soil mechanics -- Periodicals; Rock mechanics -- Periodicals; Environmental engineering -- Periodicals; 624.151
• DOI: 10.1080/17486025.2021.2008518
• Languages: English
• ISSNs: 1748-6025

Abstract

ABSTRACT: The effect of overburden stress on the rock mass deformation modulus is a known issue. However, the effect of overburden stress has been studied less with empirical methods due to the lack of appropriate data. In this study, it is aimed to investigate the effect of overburden stress on rock mass deformation modulus using artificial neural network (ANN). Four ANN models have been developed in accordance with the purpose of the study. Two of these models do not contain the overburden stress parameter, but the other two models contain the overburden stress parameter. The prediction performance of the models containing the overburden stress parameter was obtained drastically higher than the others. In other words, the value account for (VAF) and root-mean-square error (RMSE) indices of the model having the inputs of rock mass rating (RMR) and elasticity modulus of intact rock (Ei ) are 73.3% and 462, respectively, while those of the model having the inputs of RMR, Ei and overburden stress are 90% and 265. The other models developed in the present study yielded similar results. Consequently, with the ANN models developed in this study, the effect of overburden stress on Em is revealed, clearly.