Shear strength estimation of plastic clays with statistical and neural approaches

• Published in: Building and environment 2008-05, Vol.43 (5), p.849-860
• Main Authors: Goktepe, A. Burak, Altun, Selim, Altintas, Gokhan, Tan, Ozcan
• Format: Article
• Language: English
• Subjects: Applied sciences; Buildings. Public works; Computation methods. Tables. Charts; Exact sciences and technology; Geotechnics; Soil investigations. Testing; Structural analysis. Stresses
• ISSN: 0360-1323; 1873-684X
• DOI: 10.1016/j.buildenv.2007.01.022

Because shear strength parameters highly influence the bearing capacity of soils, several researchers have carried out large number of experimental and theoretical studies aimed at understanding soil strength behaviors. Within this context, the determination of correlations between soil index properties and shear strength parameters for specific soil types is possible. The aim of this study is to observe the performance of statistical and artificial neural network (ANN)-based methods on establishing correlations between index properties and shear strength parameters of normally consolidated plastic clays. To collect modeling data, consolidated-undrained triaxial tests were performed on normally consolidated plastic clays obtained from the same region. Additionally, detailed statistical analyses were conducted on the test data. Results indicate that the ANN-based model is superior in determining the relationships between index properties and shear strength parameters. However, in order to get appropriate outcomes, specific care must be dedicated when applying ANN-based correlation models.