Probabilistic Assessment of Soil Liquefaction Potential and Mitigation

It has been well observed and reported that much of the great losses in past earthquakes, such as the 2011 Tohoku earthquake and the 2010-2011 Canterbury earthquake, were attributed to soil liquefaction and the associated ground deformation. Thus, any relevant research that contributes to the worldwide efforts to assess and mitigate liquefaction hazards is considered timely and worthwhile. This dissertation is aimed at addressing two aspects of liquefaction research: (1) improving the existing probabilistic methods for both location-specific and areal liquefaction potential evaluation, (2) creating visualization-based procedure for assessing the effectiveness of dynamic compaction in the liquefaction hazards mitigation. Both are deemed timely contributions to the course of earthquake hazard mitigation efforts by the engineering communities, which are the main objectives of the research. The dissertation research consists of three separate but related efforts that as a whole address the two main objectives of this research. The first part, Predicting liquefaction probability based on shear wave velocity: an update , was intended to improve the existing liquefaction evaluation method using shear wave velocity (Vs). The liquefaction evaluation models using Vs were calibrated based on the expanded Vs-based database was created. In this work, the scientific merits of various generalized linear regression models were investigated. Based on the findings of this investigation, the optimal models were recommended for the evaluation of location-specific liquefaction probability. In the second part of the dissertation research, concerning the Random field-based regional liquefaction hazard mapping — data inference and model verification using a synthetic digital soil field , the focus was on the areal or regional evaluation of liquefaction potential. Although the random field has been applied to many geotechnical problems, including liquefaction evaluation, abundant field data for assessing various issues of random field modeling, such as the accuracy and the computational demand, are lacking. To this end, an extremely detailed three-dimensional synthetic digital soil field was created, which enabled an extensive data inference and model calibration using the random field theories. This part of the dissertation work was more on fundamental scientific exploration. Nevertheless, it set the foundation for establishing the random field-based visualization procedure for liquefaction mitigation problem in the third part of this dissertation work. In the third and last part of the dissertation work: Mitigation of liquefaction hazard by dynamic compaction — a random field perspective , the effectiveness of dynamic compaction (DC) in the mitigation of liquefaction hazards was assessed from a random field perspective. The traditional assessment of this effectiveness was through in situ tests before and after DC, and the effectiveness of such approach depends on whether the one-to-one and side-by-side field tests before and after DC are available. In reality, such ideal situation almost always does not exist due to the construction practicality in the operation of DC. The random field modeling removed such need for the one-to-one and side-by-side field tests before and after DC. In this part, a random field based visualization procedure was created so that the liquefaction potential at the entire project site before and after DC could be clearly compared. The random field based visualization procedure was demonstrated as a practical tool by which the effect of DC could be easily communicated between the engineers and their clients. The scientific endeavor in the creation of a random field based visualization procedure to help solve a practical problem was deemed significant. In summary, the three parts of this dissertation work as a whole have achieved the two main objectives of the research regarding the liquefaction potential evaluation and the liquefaction mitigation. The scientific merits through these three parts of dissertation work have been demonstrated