Effective stress based constitutive modelling and assessment of seismic pile-soil interaction in liquefiable soils


Thesis Type: Doctorate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Civil Engineering, Turkey

Approval Date: 2014

Student: SEVİNÇ ÜNSAL ORAL

Supervisor: KEMAL ÖNDER ÇETİN

Abstract:

The assessment of liquefaction-induced deformations of foundation soils located in seismically active regions has been a major concern for geotechnical earthquake engineers. Inspite of the existing efforts, the prediction of these deformations has remained a "soft" area of practice. Inspired by this gap, a new fully coupled, two-dimensional, effective stress-based, nonlinear and simplified constitutive model, referred to as METUSAND, is developed. As a part of the model development efforts, a “C++” subroutine was implemented in commercially available software FLAC based on semi-empirical cyclic straining and excess pore water pressure assessment models of Cetin et al., Cetin and Bilge and Shamoto et al. For verification purposes, well known liquefaction-induced ground deformation case history sites of Wildlife Site, Imperial Valley, California, shaken by 1987 Superstition Hills earthquake and Port Island Array, Kobe, shaken by 1995 Hyogo-ken Nanbu earthquake are re-assessed by using METUSAND. The analysis results confirmed that the proposed semi-empirical constitutive model METUSAND, can reliably predict liquefaction triggering and post liquefaction straining responses. Then, for the purpose of assessing lateral seismic deformation behavior of a single pile buried in liquefiable soils, series of numerical simulations were performed. On the basis of these numerical simulation results, and inspired by NAVFAC and Duncan Characteristic Load Method, semiempirical models to assess seismic deformation performance of piles buried in liquefiable soils were developed. Again, these recommended models were validated with actual well documented centrifuge test results. The validation studies confirmed that the proposed framework can predict pile deformations accurately within a precision of a factor of maximum two.