Prediction of maximum shear modulus (Gmax) of granular soil using empirical, neural network and adaptive neuro fuzzy inference system models

Maximum shear modulus (Gmax or G0) is an important soil property useful for many engineering applications, such as the analysis of soil-structure interactions, soil stability, liquefaction evaluation, ground deformation and performance of seismic design. In the current study, bender element (BE) tests are used to evaluate the effect of the void ratio, effective confining pressure, grading characteristics (D50, Cu and Cc), anisotropic consolidation and initial fabric anisotropy produced during specimen preparation on the Gmax of sand-gravel mixtures. Based on the tests results, an empirical equation is proposed to predict Gmax in granular soils, evaluated by the experimental data. The artificial neural network (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS) models were also applied. Coefficient of determination (R2) and Root Mean Square Error (RMSE) between predicted and measured values of Gmax were calculated for the empirical equation, ANN and ANFIS. The results indicate that all methods accuracy is high; however, ANFIS achieves the highest accuracy amongst the presented methods.

Automated summary

The maximum shear modulus (Gmax) of a soil mixture is influenced by factors such as void ratio, effective confining pressure, grading characteristics, consolidation, and initial fabric anisotropy. This study proposes an empirical equation to predict Gmax in granular soils based on bender element tests and evaluates the accuracy using artificial neural network (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS) models. The results show that all methods have high accuracy, with ANFIS achieving the highest accuracy.