In geotechnical earthquake engineering, evaluating the risk of liquefaction hazards is an important issue for modern seismic engineering design. In Taiwan seismic design codes of ordinary buildings, there are three indicators suggested to quantify the severity of ground manifestation or the damage of low-rise buildings due to liquefaction: the liquefaction potential indicator (LPI) proposed by Iwasaki et al. (1982), the thickness of non-liquefied surface layers (H1) proposed by Ishihara (1985), and the post-liquefaction settlements (SL) proposed by Ishihara and Yoshimine (1992). Associated with these three indicators, the criteria and principles for liquefaction risk evaluation recommended by the corresponding originators are also prevalent to be adopted as indicators for the occurrence of liquefaction in seismic design in Taiwan practice; however, these criteria were generally derived from the investigation and observations from Japan case histories of liquefaction along with calculati on procedures developed in Japan, such that their applicability to Taiwan sites and assessing methods in Taiwan seismic design codes remains unknown. To determine the indicator of occurrence of ground manifestation of liquefaction for the practice in Taiwan, this paper exploits the liquefied and non-liquefied case histories of the 1999 Chi-Chi earthquake and those of the 2016 Meinong earthquake in Taiwan and evaluates their LPI, SL, and H1 values by 4 SPT-based simplified methods which are suggested in Taiwan seismic design codes for the ordinary buildings. Based on the combinations of these values, several prediction models for the surface manifestation of liquefaction can be determined using statistical approaches. These prediction models are compared and discussed in this paper, which is believed as a reference to develop future seismic design recommendations on strategies for liquefaction prevention.

Simplified liquefaction assessment procedures have been widely used to estimate the severity of liquefaction under earthquake loading, which is often confirmed by the presence of sand boils. Back analyses of case studies have shown that the simplified assessment procedures can overestimate or underestimate the liquefaction potential of a deposit (i.e. false positives and false negatives respectively). The occurrence and severity of sand boils are highly dependent on the soil permeability and the hydro-mechanical interaction of cross-layers (system response effects), which the simplified liquefaction assessment procedures do not account for. Additionally, the variation of hydraulic conditions during and after the earthquake shaking significantly affects the evolution of sand boils. In this study, the layered system effects on liquefiable deposits are examined through dynamic nonlinear effective stress analysis. Scenarios where a liquefiable layer is interbedded within materials of distinct hydro-mechanical characteristics are examined parametrically employing the fully coupled (u-p) finite element software PLAXIS. The influence of the presence and the characteristics of non-liquefied or low-permeability layers surrounding the liquefiable layer on the development of sand boils is investigated.

This paper focuses on the evaluation of effects of partial saturation on the liquefaction response of two free-field level-ground deposits from Christchurch (New Zealand). The first deposit is composed of vertically continuous liquefiable soils with a low-resistance critical zone at shallow depth and is typical of sites that manifested moderate-to-severe liquefaction at the ground surface in several events during the 2010-2011 Canterbury earthquake sequence. The second deposit consists of liquefiable soils of low liquefaction resistance interbedded with non-liquefiable layers, and is typical of sites that did not manifest liquefaction in any of the 2010-2011 seismic events. High-resolution measurements of compression wave velocity (V_p) have indicated the existence of a partially saturated zone at shallow depth below the groundwater table in both deposits, though with somewhat different characteristics. We assess the performance of the two deposits for the 22 February 2011 Christchurch earthquake using simplified liquefaction triggering analysis as well as advanced nonlinear dynamic analysis. Partial saturation effects are considered in both types of analyses by “correcting” the liquefaction resistance of the partially saturated soils on the basis of an empirical V_p-based relationship. The analyses indicate that partial saturation contributes to the formation of a specific sequence of system-response mechanisms that collectively act to mitigate liquefaction manifestation in the case of the interbedded deposit. In the case of the vertically continuous deposit, however, the mitigating effect of partial saturation is counteracted by system-response mechanisms that intensify the effects of liquefaction. The results highlight the importance of considering the effects of partial saturation in the context of the overall system response of liquefying deposits and consequent liquefaction manifestation.

The importance of soil saturation condition in relation to liquefaction occurrence emphasizes the necessity for a reliable tool to accurately monitor soil moisture content during seismic events. The Soil Moisture Active Passive (SMAP) satellite provides near-real time measurements of surface and root zone soil moisture globally. Common proxies used to assess soil saturation in liquefaction analysis, such as water table depth patterns, historical mean annual precipitation measurements, and topographic conditions, may not be as effective. To address this, this paper introduces the use of satellite-based soil moisture data to enhance the understanding of the relationship between saturation conditions and liquefaction events. Established geospatial explanatory variables, along with new SMAP-based soil moisture parameters, were employed to develop a recently proposed global liquefaction model called the Soil Moisture-based Global Liquefaction Model (SMGLM), which was compared to an existing global liquefaction model. However, limitation of SMGLM arising from the coarse spatial resolution of remotely sensed soil moisture data highlights the need for updated SMGLM using soil moisture data with finer resolution. With the continuous advancement of earth observing satellites, the findings of this study could serve as a foundation for developing fully satellite-based models that utilize high-resolution, near-real time soil moisture data to identify liquefied sites. Incorporating near-real time ground-truth-based saturation condition is a crucial step towards addressing potential overestimation or underestimation of liquefaction likelihood that may result from using proxies for soil saturation in existing models.

Liquefaction susceptibility assessments are heavily influenced by cyclic resistance curves derived through undrained element testing. The undrained experimental condition is based on the hypothesis that water flow and the resulting void redistribution is negligible within the time-frame of an earthquake. However, this hypothesis has come into question following recent analytical and experimental evidence. Depending on soil properties and stratification, localised, co-seismic volumetric strains can either facilitate or hinder the triggering of liquefaction. This paper is concerned with cases where partial drainage results in localised volumetric contraction. This condition could take hold close to the interface of a liquefiable layer with an underlying layer of lower permeability, or close to the interface of a liquefiable layer with an overlying layer of higher permeability. The aim of this work is to examine the effect of volumetric contraction on cyclic resistance. For the sake of simplicity, only constant rates of volumetric contraction are examined. The results presented are obtained using load-controlled cyclic triaxial element tests, performed on specimens of medium-dense Hostun sand. The undrained cyclic resistance curve is initially established. Subsequently, cyclic triaxial tests under constant volumetric contraction rates are presented. They reveal that volumetric strains can be of paramount importance, as even very small rates of volumetric contraction can have a dramatic effect on the number of loading cycles required to reach liquefaction.

Currently, shear wave measurement is an effective method for assessing the strength of foundation soils and is widely used in field engineering and laboratory tests. Bender element test has been widely placed as an indoor shear wave testing device to measure strength parameters such as shear modulus. However, the propagation characteristics of shear waves in a medium highly depend on the stress state of the medium itself and its changes. As the propagation medium of shear waves, the effective stress state and the skeleton structure of soil are inevitably affected when subjected to loading history. Therefore, in order to investigate the propagation characteristics of shear waves under multiple loading disturbance in the previous period, a series of indoor dynamic triaxial tests and bender element tests were carried out using Toyoura sand, and different stress paths were applied to the sand to investigate the propagation behavior of shear waves under the action of different dynamic loading stress paths and isotropic stress conditions.