Abstract
Unsatisfactory performance of pile supported structures in liquefiable areas (ranging from tilting/settlement to complete collapse) is still observed after most major earthquakes. As a result, further research is required in this subject. This thesis therefore aims to study the response of pile supported structures during seismic liquefaction. The ground liquefies progressively in a top down fashion when the soil transform from solid material to liquid-like material. This is referred to as transient behaviour (from no-liquefaction to full liquefaction state) and is particularly focused in this work. In practice, piles are usually analysed as laterally loaded beams using Beam on Nonlinear Winkler Foundation model where earthquake loading is applied in a pseudo-static way. Therefore, this study reviews methods of analysis of laterally loaded pile. Six different field case records were analysed using different approaches and the results were compared. Large scale shake table experiments were also conducted consisting of four pile models (two single piles and two pile groups of 2×2) placed in a rigid soil container with energy absorbing boundaries. Redhill-110 sand was used and earthquake motions were applied to liquefy the soil. It was observed that the bending moment along the piles changed with the progression of liquefaction and the maximum bending moment occurred in the transient phase. It was also observed that the time taken to reach liquefaction may affect the amplification of the bending moment. Design of piles requires soil parameters and as a result, a series of multi-stage soil element tests were carried out on four different types of sands; Redhill-110 sand, Japanese silica sand No. 8, Assam sand, and Ganga sand where the sands were first liquefied and then tests were carried out to obtain stress-strain of liquefied sand (post-liquefaction). The results showed that the post liquefaction behaviour of sand depends on the soil relative density. Furthermore, the results from the Redhill-110 sand were used to back analyse the shake table test results. Finally, a method has been proposed to incorporate transient behaviour of pile in liquefiable soils, based on an assessment of the estimated dynamics amplification factors in the shake table tests. Keywords: Dynamic soil-pile interaction, Liquefaction, Shake table test, multi-stage soil element test, transient dynamics, dynamic amplification factors.