Thesis Type: Doctorate
Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Civil Engineering, Turkey
Approval Date: 2010
Student: HABİB TOLGA BİLGE
Supervisor: KEMAL ÖNDER ÇETİNAbstract:
Although silt and clay mixtures were mostly considered to be resistant to cyclic loading due to cohesional components of their shear strength, ground failure case histories compiled from fine grained soil profiles after recent earthquakes (e.g. 1994 Northridge, 1999 Adapazarı, 1999 Chi-Chi) revealed that the responses of low plasticity silt and clay mixtures are also critical under cyclic loading. Consequently, understanding the cyclic response of these soils has become a recent challenge in geotechnical earthquake engineering practice. While most of the current attention focuses on the assessment of liquefaction susceptibility of fine-grained soils, it is believed that cyclic strain and strength assessments of silt and clay mixtures need to be also studied as part of complementary critical research components. Inspired by these gaps, a comprehensive laboratory testing program was designed. As part of the laboratory testing program 64 stress-controlled cyclic triaxial tests, 59 static strain-controlled consolidated undrained triaxial tests, 17 oedometer, 196 soil classification tests including sieve analyses, hydrometer, and consistency tests were performed. Additionally 116 cyclic triaxial test results were compiled from available literature. Based on this data probability-based semi-empirical models were developed to assess liquefaction susceptibility and cyclic-induced shear strength loss, cyclically-induced maximum shear, post-cyclic volumetric and residual shear strains of silt and clay mixtures. Performance comparisons of the proposed model alternatives were studied, and it is shown that the proposed models follow an unbiased trend and produce superior predictions of the observed laboratory test response. Superiority of the proposed alternative models was proven by relatively smaller model errors (residuals).