Abstract

The presence of a non-liquefied crust overlying a liquefied layer has been found to have a significant effect on sand ejecta and surficial liquefaction damage, as observed by Ishihara (1985). Following the 2010-2011 Canterbury seismic sequence in New Zealand, almost no foundation deformation occurred in areas with liquefaction susceptible soils overlain by at least a 3 meter thick crust [1]. In contrast, the 2012 Emilia-Romagna earthquake in Italy provided surface evidence of liquefaction despite 6+ meter thick crusts. Several researchers have proposed liquefaction prediction models [e.g. LPI [2], Ishihara Curves [3], LSN [4], LPIISH [5] and the methods proposed by Towhata, et al. [6] and Hutabarat and Bray [7]]. However, few methods consider the properties of the non-liquefiable caps. Numerous soil properties of the non-liquefiable caps from 2000+ Cone Penetration Tests (CPTs) were calculated and analyzed to consider their effects on liquefaction surface ejecta manifestations. These properties were considered in conjunction with existing liquefaction prediction models for the 2012 Emilia-Romagna earthquakes, and a combination of soil properties was found to successfully improve surface ejecta manifestation predictions. A new method, the Liquefaction Ejecta Prediction Model (LEPM), was created using these data. LEPM combines LPI with this combination of predictive soil properties to develop a model with more predictive power than any of the existing methods tested in this study on this set of case histories. LEPM was created as a logistic regression model to assess liquefaction hazard from a probabilistic approach.

Degree

MS

College and Department

Ira A. Fulton College of Engineering; Civil and Environmental Engineering

Rights

https://lib.byu.edu/about/copyright/

Date Submitted

2024-12-17

Document Type

Thesis

Keywords

Liquefaction; Cone penetration tests; Surface manifestation; Logistic regression

Language

english

Download

Included in

Engineering Commons

Share

COinS