Title: Soft computing and statistical analysis methods to forecast the uniaxial compressive strength of marl rocks

Authors: Krikar M. Gharrib Noori; Younis Mustafa Ali Alshkane; Kamal Ahmad Rashed

Addresses: College of Engineering, University of Salahaddin, Erbil, Iraq ' Faculty of Engineering, University of Sulaimani, Sulaimani, Iraq ' Faculty of Engineering, University of Sulaimani, Sulaimani, Iraq

Abstract: The purpose of this study is to forecast the uniaxial compressive strength (UCS) of marl rocks, which is a crucial indicator in understanding the rock strength and evaluating their suitability for engineering applications. Simple linear regression (SLR) and artificial neural networks (ANN) are two examples of soft computing techniques used in the research. The study places a focus on data normality, model quality, and accuracy using information from 119 samples of marl rock in Iraq-Kurdistan. With an R-squared value of 0.76 for UCS prediction, the point load index Is₍₅₀₎ parameter of the SLR models produced the model (M4) that was most appropriate. However, the ANN-M4 model performed the best, producing the most precise UCS predictions. Montmorillonite clay minerals were discovered during mineralogical and geochemical analyses using SEM, EDX, XRD, and XRF. This information explained why the rock was susceptible to changes in moisture content and potential disintegration.

Keywords: rock engineering parameters; statistical analysis SLR; soft computing ANN-MLP; mineralogical investigation; marl rock type.

DOI: 10.1504/IJMRI.2025.144875

International Journal of Masonry Research and Innovation, 2025 Vol.10 No.2, pp.125 - 144

Received: 22 Sep 2023
Accepted: 06 Nov 2023
Published online: 06 Mar 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article