| dc.description.abstract | Since Seed and Idriss introduced a simplified procedure for evaluating soil liquefaction potential in 1971 [1], scholars have developed various liquefaction assessment methods using local soil liquefaction databases. This research project aims to evaluate whether the results from different methods are statistically equal based on statistical hypothesis testing under three seismic conditions: small earthquake (Mw = 6.5, PGA = 0.1 g), moderate earthquake (Mw = 7.0, PGA = 0.3 g), and large earthquake (Mw = 7.5, PGA = 0.5 g). Liquefaction safety factors were computed for of the collected soil layers, based on which this study was proceeded using hypothesis testing and other analytics.
Considering the influence of random sampling, this study employs analysis of variance and hypothesis testing to analyze the statistics of liquefaction safety factors. The results indicate that the JRA method is relatively conservative compared to the other four methods under small and moderate earthquake scenarios. However, under large earthquake scenarios, the liquefaction safety factors from HBF, NCEER, and B&I methods can be considered equal from the perspective of statistical hypothesis testing. On the other hand, the AIJ method yields higher safety factors across different earthquake scenarios. Finally, when comparing the AIJ method with the most conservative method under each seismic scenario, the differences are approximately 30-33% (small earthquake), 24-27% (moderate earthquake), and 10-25% (large earthquake) of the average safety factors obtained by the AIJ method. | en_US |