Can geotechnical liquefaction indices serve as predictors of foundation settlement?

Geotechnical liquefaction indices, such as the liquefaction potential index, are commonly used as proxies for the risk of liquefaction-induced damage at site or regional scales. However, these indices were developed based on surficial manifestations of soil liquefaction in the free field, and, as such, they have been shown to correlate better with land damage than foundation damage. This study evaluates the ability of three geotechnical liquefaction indices to predict foundation settlement on liquefiable soils, as compared to both conventional ground motion intensity measures (IMs) and the term for site and ground motion effects in a probabilistic model specifically developed for foundation settlement. A new metric for the predictive ability of these measures, skill, is proposed to quantify the total uncertainty in settlement predictions using a given measure. The Ishihara-inspired liquefaction potential index is found to be the optimum index, and cumulative absolute velocity (CAV) as predicted on outcropping rock is found to be the optimum IM. However, although both measures are regionally applicable, neither outperforms the site term from the probabilistic settlement model, which was developed using the same numerical database used in this study.

Automated summary

The study compared the predictive ability of three geotechnical liquefaction indices and conventional ground motion intensity measures in predicting foundation settlement on liquefiable soils. It found that the Ishihara-inspired liquefaction potential index and cumulative absolute velocity (CAV) on outcropping rock were the optimal indices to use. However, neither of these measures outperformed the site term from the probabilistic settlement model developed using the same database.