Seismic response of end-bearing fibre-reinforced polymer (FRP) piles in cohesionless soils

Abstract

This study focuses on investigating seismic performance of hollow fibre-reinforced polymer (FRP) piles in liquefiable sand deposits using shaking table tests. A series of soil-foundation models were prepared in a laminar shear box with a dimension of 1.0 m\u2009×\u20091.0 m and a depth of 1.0 m following scaling relationships that recognize the dynamic and nonlinear nature of soil–pile systems. Four glass fibre-reinforced polymer (GFRP) piles and four carbon fibre-reinforced polymer (CFRP) piles were manufactured and embedded as end-bearing piles within the soil and tested under different seismic input motions. A group of four aluminium piles were also employed and tested under similar conditions to compare their response with FRP piles. The model soil was prepared following pluviation technique inside the laminar shear container while a thin flexible latex membrane was stretched inside the container to contain the soil and water. The cycles of shaking tests were carried out using ground motions from the 1995 Kobe Earthquake and the 2010 Central Canada Earthquake. Results of strong excitation (K20) indicated lower deviation of foundation to free field motions in the frequency range of 10–20 Hz and, hence, lower amplification of input motions for FRP piles (i.e. response spectra ratio of 1.45 to 2.20 times) due to higher flexibility of hollow FRP piles compared to conventional piles. Among the FRP piles, the glass FRP showed better performance than the carbon FRP piles in terms of soil–pile interaction and foundation input motions. Pile caps have also experienced smaller amplification for GFRP piles as a result of their higher flexibility along the pile shaft with peak acceleration of 0.3 g. Results indicated that hollow FRP pile can be a suitable alternative in liquefiable soils because of their favourable material characteristics in seismic prone area.