A Comprehensive Study for Assessing Parameters Influencing Tensile Strength Behaviour of Fine-Grained and Coarse-Grained Soils

Development of tensile cracks in earthen structures is majorly controlled by tensile strength, which is affected by a number of parameters. The influence of different soil properties and their relative dominance on tensile strength characteristics of both fine- and coarse-grained soils has been brought out by systematic investigations using Artificial Neural Network (ANN) technique, in the present study. A few important soil parameters considered are degree of saturation, dry density, porosity, suction, plasticity index, fines content, relative density, average particle diameter, and sand content. Several ANN models, independently for fine- and coarse-grained soils, consisting of minimum one to a maximum six input parameters have been developed. Results demonstrate that influence of parameter(s) is distinct on tensile strength of fine- and coarse-grained soils. Results also clearly manifested usefulness of ANN tool to discern which parameter(s) could largely govern the tensile behaviour of soils. Amongst many parameters, fines content, sand content and average particle diameter are found prominently influencing and ascribed as must to include parameters for determining the tensile strength of soils respectively. It has been noticed that the significance of these grain size parameters is analogous to microstructural hydromechanics of soil and, thereby, accounted for universal tensile behaviour of soils.