Belonging to soils of Adapazari; Investigation of the relationships between physical, mechanical and dynamic properties

To determine the physical, mechanical and dynamic properties of soils is very important for geotechnical engineering applications. In the analysis of many geotechnical engineering projects such as bearing capacity of shallow and deep foundations, slope stability and deep excavations, shear strength parameters of soils must be correctly estimated. Many researchers have been performed statistical studies on relationships between the physical, mechanical and dynamic properties of soils. Recently, Artificial Neural Networks (ANN) which is capable of higher estimates compared to classical regression analysis has begun to take place frequently in geotechnical engineering practice. In this study, the relationships between the physical, mechanical and dynamic properties of fine-grained soils have been investigated with multiple regression analysis and artificial neural network methods. For this purpose, experimental results belonging to the undisturbed samples obtained from drilling studies applied in Adapazari city borders are used. Firstly, the presence of relationships between the physical, mechanical and dynamic properties of the undisturbed samples were investigated with regression analysis. In addition, the predictability of shear strength from the index properties was evaluated by using ANN. Consequently, ANN has provided acceptable results on determining the shear strength of fine-grained soils from index properties considering the local soil conditions in the study area.