Prediction of Small-Strain Properties of Dry Sand Using Curve Fitting and Machine Learning Models

Department of Aerospace Engineering, Indian Institute of Science, Bengaluru, 560012, Karnataka, India

^* e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.

Published online: 1 December 2025

Abstract

This study investigates the prediction of small-strain modulus and Poisson’s ratio of dry sand using both curve-fitting techniques and machine learning models. The stress-strain data utilized for this work were generated through YADE DEM triaxial simulations, which were first validated against experimental data to ensure accuracy and reliability. The small-strain modulus and Poisson’s ratio were modeled as functions of confining pressure and initial void ratio. Curve-fitting techniques were employed to develop constitutive models describing these relationships. Additionally, various machine learning models were used to predict the smallstrain modulus and Poisson’s ratio, capturing complex nonlinear dependencies. A systematic comparison of the curve-fitting and machine learning approaches was conducted to evaluate their performance in terms of accuracy and generalizability. The findings underscore the potential of machine learning to improve predictive accuracy, offering valuable insights for soil mechanics and geotechnical engineering applications.