Artificial Neural Network Approach Based Indirect Estimation of Shear Strength Parameters of Soil

The shear strength parameters of soil such as cohesion (C) and angle of internal friction (φ )govern the bearing capacity of shallow or deep foundations, the stability of the hill slopes as well as slopes of dams and embankments, design of retaining structures etc. in most of the civil engineering practices. In this study, artificial neural network (ANN) with the feed-forward back-propagation algorithm is proposed for predicting the shear strength parameters of the soils. To develop this ANN based model, different soil parameters such as gravel % (GP), sand % (SP), silt % (STP), clay % (CP), dry density (DD) and plasticity index (PI) obtained through laboratory tests for soil samples from different parts of India, have been used as input parameters to predict the shear parameters of the soil. In case of data sampling using fuzzy clustering method, the neural network with architecture 6-9-2 produced best correlation coefficient(R) and RMSE values for both training and testing. The R values for training and testing in this case are 0.73 and 0.86 respectively for prediction of both C and φ . While for random data sampling, the neural network 6-3-2 produced best results in terms of R and RMSE values. The R values in this case for training and testing are 0.81 and 0.86 respectively for prediction of both C and φ . It can be inferred from the results that indirect estimation of shear strength parameters of soils is feasible with reasonable accuracy using ANN technique. INTRODUCTION In most of the civil engineering practices the shear strength parameters such as cohesion and angle of internal friction play a big role in depicting the hill slope stability, dam’s inclination, bearing capacity for foundations such as deep or shallow, embankments and retaining structures design. Based on the particle size or on the distribution of particles the soil can be actually defined. For geotechnical purposes this soil can be divided into two categories. The first comes as cohesive soil and the second as cohesion less soil. If soil has a diameter lower than 0.067 mm then it is termed as cohesive soil. This soil is mostly composed of clay and silt. Where as, soil having diameter greater than 0.067mm is non cohesive soil. The main composition are sand and gravels. Mohr Coulomb theory represents the shear strength of the geotechnical materials. The shear soils strength is directly proportional to the stress applied through shear strength components such as cohesion and internal friction angle. Fig 1 shows the slope and intercept formed in the tangent envelope is represented by the Mohr Coulomb failure envelope, where the slope represents the angle of internal friction and the intercept formed is the cohesion. The shear strength parameters calculated from the laboratory is quite common and expensive in nature. It is always not feasible to conduct tests for every situation regardless of the conditions. One