Prediction of Bearing Capacity of a Footing Resting on Geo-Synthetic-Reinforced Soil Wall Using Artificial Neural Network

Increased recent application of geo-synthetic-reinforced soil (GRS) walls as bridge abutments to support bridge beams over shallow foundations is pervasive in place of deep foundations. Understanding the behavior of footing resting on the backfill of the GRS wall is necessary and finding out the bearing capacity of the footing is essential. Many researchers have calculated the bearing capacity of the footing resting on GRS walls by using numerical analysis through various softwares. In the present study, numerical analysis is performed to estimate the effects of various factors, namely embedment depth of footing, angle of internal friction, offset distance of footing, width of footing, and length of reinforcement on bearing capacity. Consequently, an artificial neural network (ANN) is applied to predict the bearing capacity of the footing. For this, 190 data points collected from previous research articles and others processed in PLAXIS 2D software are used in the present analysis. A model equation for the determination of the ultimate bearing capacity of the footing resting on the GRS wall has been developed from the best-fit ANN model. Finally, sensitivity analysis was performed to determine the order of importance of input parameters on the output parameter.