Fuzzy-Based Model for Predicting Strength of Geogrid-Reinforced Subgrade Soil with Optimal Depth of Geogrid Reinforcement

Abstract

A fuzzy logic (FL)–based modeling approach is employed for geogrid-reinforced subgrade soil of unpaved roads. A review of the literature reveals that fuzzy logic has not been used for predicting the behavior of geogrid-reinforced subgrade. This paper presents FL-based two models with fuzzy Triangular and Gaussian membership functions for input and output variables. It consists of eight input parameters/factors, namely, reinforced/unreinforced section, depth of reinforcement, liquid limit, plastic limit, plasticity index, optimum moisture content, maximum dry unit weight, and soaked/unsoaked condition, and California bearing ratio (CBR) as an output parameter. The fuzzy rules are deduced from the experimental data. The laboratory CBR tests were performed on the subgrade soil reinforced with geogrid. The precision of models was examined by comparing the predicted CBR values with the experimental CBR values for Triangular and Gaussian membership functions. The sensitivity analysis reflects a set of dominant parameters. The results indicated a significant improvement in the CBR value of geogrid-reinforced subgrade soil due to the inclusion of geogrid. The range for optimal depth of geogrid reinforcement is found to be 36 to 60% of the thickness of the soil layer. The potentialities of FL were found to be satisfactory.