Evaluation of lateral spreading due to Canterbury earthquakes using effective stress analysis

Following the September 4th 2010 M w 7.1 and February 22nd 2011 M w 6.2 Earthquakes in Canterbury, New Zealand, substantial lateral spreading was observed along the Avon and Heathcote rivers. Detailed field measurements following the events provides a benchmark to evaluate the efficacy of lateral spreading prediction models for realistic conditions. This paper presents the development of 2D effective stress finite difference models in FLAC of seven transects along the rivers, and compares the results to observed displacements for the two major seismic events. The liquefiable soil was modelled with the PM4Sand model, and an automated calibration procedure was developed to determine the model inputs based on CPT results. Each transect was subjected to 20 dynamic analyses, where two different ground motions were used for each event and five scaling factors were applied. The displacements during post-shaking pore pressure dissipation were also quantified. Overall the simulations show reasonably consistent ground displacements at the channel free-face for both events for all transects. Noting that only for one case the best estimate of simulation was outside the range of 1:1.5 and 1.5:1 of the best estimate from field measurements, which was also the only transect-event combination for which the overlap between simulation and measured values was outside the considered uncertainty bounds. The results of the numerical simulations highlight the importance of both geometric (notably the channel width), soil profile characteristics and soil properties as key factors that influence lateral spreading displacements, as well as input ground motion intensity. The simulation results provide confidence that detailed finite difference analysis can reasonably capture the expected level of ground movement during a lateral spreading event. • Simulations of lateral spreading case histories that consider multiple earthquakes as well as small and large magnitude lateral displacements. • Quantification of post-shaking lateral displacement. • Quantification of the role of calibration, ground motion selection, and specific modelling decisions on simulated lateral displacements. • Impact of the width of the river channel on lateral displacements.