Numerical modelling of rammed aggregate piers (RAP) in liquefiable soil

Liquefaction poses a significant risk to the built environment, and as a result, ground improvement (GI) techniques are commonly used to mitigate this risk. Liquefaction mitigation strategies continually evolve, and several relatively new techniques, such as Rammed Aggregate Piers (R) (RAP), have shown promise. While the densification mechanism associated with many of the GI techniques is generally well known, other mechanisms such as reinforcement, lateral stress increase, and improved drainage that are thought to enhance liquefaction mitigation are still not completely understood. Moreover, field performance data for these GI schemes during actual earthquakes is limited. To fill this gap and evaluate the performance of RAP, this study presents the first numerical model on a specific GI technique that uses detailed site characterization, large-scale field test data, and post-earthquake field performance observations to calibrate and qualitatively validate the model. The in-situ characterization and full-scale field test data collected from the Ground Improvement Programme (GIP) performed following the 2010-2011 Canterbury Earthquake Sequence (CES) in New Zealand are used in this study. A set of fully-coupled hydro-mechanical finite difference (FD) models are performed in natural and reinforced conditions. For the unimproved soil profile, the results predict shear strains and zones of high excess pore water pressures reasonably well. For the profile reinforced with RAP, the analyses indicate that the stiffness properties of the RAP have the greatest influence on the reduction of generated shear strains in the soil profile. Finally, the calibrated models are subjected to a set of ground motions with different intensities to assess the efficacy of the RAP under different loading conditions.