AI Summary & Key Takeaways
AI InsightThe highly accurate prediction performance of the proposed model demonstrated the availability of modern methods for the estimation of soil parameters.
📌 3 Key Takeaways
Abstract
Determination of soil deformation and strength properties is of great importance in geotechnical design. The disturbance that occurred during the sampling from the field affects the results obtained by the laboratory tests. Therefore, the lack of representation of the field conditions and in-situ soil features with laboratory tests leads the designers to carry out field tests. Standard penetration test (SPT) and the Menard pressuremeter test (PMT) are the most widely used geotechnical field tests in which the results are utilized to investigate soil properties and lateral deformation characteristics at a specified depth. In this study, a total of 102 data of sandy and clayey soils belonging to the geotechnical investigations carried out in Istanbul were compiled. The regression analysis between the corrected SPT blow count N60, pressuremeter modulus EPMT, limit pressure PL and EPMT/PL ratio is performed. Empirical equations were separately developed for sandy and clayey soils between the considered parameters. The developed equations showed that there are acceptable relationships between the parameters examined in the particular dataset. Moreover, a neural network (NN) based prediction model was developed to predict EPMT and PL using the available soil data. The highly accurate prediction performance of the proposed model demonstrated the availability of modern methods for the estimation of soil parameters.