Author: Gulmustafa ŞEN, Erdal AKYOL, Mahmut FIRAT 2007
Publishing Date: 2007
Volume 22 Issue 1-2
Liquefaction is one of the major natural hazards caused by earthquakes and it can be defined as an increase of pore pressure and lost of bearing capacity of the soils because of a dynamic impact (earthquake). In order to estimate liquefaction potential, Liquefaction Potential Index (LPI) is calculated. LPI has been developed for evaluating the surface impacts of the liquefaction. Artificial Neural Networks (ANN) developed biological human brain system has been recently used for modeling of complex and nonlinear engineering problems.
In this study LPI of the Gümüşler Municipality settlement area has been calculated and factor of safety (FS) against liquefaction has been estimated by using ANN. Feed forward type of ANN is employed. The input parameters are cyclic resistance ratio for Mw=7.5 earthquakes (CRR7.5), cyclic stress ratio (CSR), depth of ground water level, depth and N values of standard penetration test. 194 borehole values have been used in training process while 123 data have been used in test procedure. Satisfactory results have been obtained.
Key Words: Artificial Neural Networks; Factor of Safety; Liquefaction Potential Index.