This paper presents the results of 1520, three-dimensional (3D), fully-coupled, effective-stress, finite-element simulations in OpenSees to evaluate the influence of a variety of intensity measures (IMs) on excess pore pressures and potential for surface ejecta in the free-field. The analyzed profiles showcase different layering, and vertical changes in the groundwater table. We investigate the peak excess pore water pressure ratio in the middle of the liquefiable layer (peak r_u) and the ejecta potential index (EPI) as the primary engineering demand parameters (EDPs) of interest. A total of 20 IMs are considered as candidates, which represent a range of characteristics in terms of amplitude, frequency content, and duration of seismic loading. Efficiency, sufficiency, and predictability are used as the criteria to identify high-performing IMs for predicting the EDPs of interest. Based on the results from this preliminary parametric study, the average pseudo-spectral acceleration over the period range from 20% to 200% of the initial, small-strain site fundamental period, S_a,avg(0.2T_so, 2T_so) is identified as a high-performing IM for peak r_u in the free-field. For EPI predictions, the high-performing IM was found to be the cumulative absolute velocity (CAV). Both IMs were on outcropping rock.

Case histories have been critically important in advancing the field of geotechnical earthquake engineering. The sequence of Canterbury, New Zealand, earthquakes in 2010-2011 produced an unparalleled quantity of high-quality liquefaction manifestation data. The data include airborne LiDAR-based measurements of ground surface elevation, aerial and ground photographs, property inspection reports, cone penetration tests, borehole logs, groundwater depth measurements, and peak ground acceleration (PGA) estimates. These data were utilized to develop the first liquefaction ejecta database. It contains 235 well-documented case histories with estimates of liquefaction ejecta-induced settlement and parameters describing ground conditions and ground motions. The settlement was estimated using the LiDAR-based and photographic-based approaches. Most of the sites in the database have thick, clean sand deposits. About one-half of these sites settled at least 50 mm due to ejecta when shaken by a M_w 6.1-equivalent PGA > 0.40 g. The severity of liquefaction ejecta at these clean sand sites tended to be underestimated by the state-of-practice liquefaction induced-damage indices. The underestimation was greater by indices that did not consider post-shaking hydraulic mechanisms. Strong shaking and liquefaction induced by the M_w 6.2 Feb 2011 earthquake formed cracks in the crust, which likely exacerbated the severity of liquefaction ejecta in the following M_w 6.2 June 2011 earthquake. The liquefaction ejecta case histories provide important insights, some of which are shared in this paper.

The research regarding the liquefaction itself started decades ago, with sand being the main material due to its highest suspected potential to be liquefied compared to other materials. More recently, as it was proved that the fines-contained material is also suspected to be liquefied, more studies, focusing on the elementary tests, have already been conducted. However, fewer studies have been conducted to investigate the relationship between the fines content of the liquefied ground and the movements of structures in the ground during earthquakes, in this case, the uplift of the buried pipe structure. In this study, therefore, a series of 1-G shaking table tests were carried out to simulate the real-field situation in a ground condition with varying non-plastic fines content and a pipe model located within it. The experiments revealed that the ground model with a higher fines content is expected to have a lower total uplift displacement than the ground with less or no fines content, which is probably due to the effect of the strength recovery of the ground.

The design of the deep shafts is generally considered to be in a weak soil against liquefaction, however the shafts which are built as a facility in nuclear power plants, are mainly built-in dense ground. In this study, we aimed to clarify the interaction between the ground and the shaft during an earthquake, which is simulated in a dense sandy soil. To achieve this goal, a series of dynamic centrifuge model tests and liquefaction analysis based on effective stress analysis by using FLIP are performed. In the analysis 14.1m deep circular shaft is modelled as buried in 10.2m thick liquefiable dense sand layer, which is covered with 3.9m of non-liquefiable dense sand layer. The results of the analysis showed that the negative excess pore water pressure, the maximum acceleration, and the maximum ground displacement of the liquefiable ground occurred simultaneously. However, the time that maximum acceleration of the structure and the maximum displacement of the ground did not always coincide. In case of dense sandy soil, we observed a small ground deformation even if the excess pore water pressure increased. As a result, since the ground deformation, which generally has a large impact on the structures, was small, the earth pressure was also small. The obtained results confirmed that the shafts in dense sandy ground performed well under dynamic effects. Hence, in design by performing numerical analysis with comprehensive soil and structure properties, economical and conservative solutions can be achieved.

Centrifugal model experiments were conducted to investigate the cause of the difference in liquefaction damage to roadways and adjacent sidewalks. A water film is formed under the roadway’s roadbed during the dissipation of excess pore water pressure after liquefaction. For thick liquefaction layers, water in the film flows under the sidewalk’s roadbed, causing an uplift of the ground surface owing to heaving of the sidewalk. In contrast, when the liquefied layer is thin, the water film formed under the roadbed is thin, and the amount of water flowing into the sidewalk is small; thus, no uplift of the sidewalk area occurred.

Soil liquefaction occurring due to earthquakes poses severe risk to both above-ground structures as well as buried structures such as pipes, manholes etc. During liquefaction, the shear strength of the soil above and around the pipeline could decrease due to build-up of excess pore pressures and, subsequently, resulting in buoyancy forces that could cause the pipelines to displace and potentially “float up” towards the ground surface. Limit-equilibrium based procedures allow for prediction of uplift occurrence, but predicting the magnitude of uplift is a complex task with a number of components such as soil type, pipe diameter (D) and burial depth (H), pipe boundary constraints, and earthquake motion contributing to this mechanism. This study utilizes a well-calibrated and validated 2D numerical model to investigate the effects of input motion characteristics on pipe uplift for a steel pipe buried in a saturated, loose, and homogeneous deposit of Fraser River sand. The commercially available finite difference FLAC software and soil constitutive model PM4Sand were utilized. The influence of input motion amplitude and duration on uplift behavior of pipe was examined and discussed.