This paper presents a new simplified method for assesing the liquefaction resistance of soils based on the cone penetration test (CPT). A relatively large database consisting of CPT measurements and field liquefaction performance observations of historical earthquakes is analyzed. This database is first used to train an artificial neural network for predicting the occurrence and nonoccurrence of liquefaction based on soil and seismic load parameters. The successfully trained and tested neural network is then used to generate a set of artificial data points that collectively define the liquefaction boundary surface, the limit state function. An empirical equation is further obtained by regression analysis to approximate the unknown limit state function. The empirical equation developed represents a deterministic method for assessing liquefaction resistance using the CPT. Based on this newly developed deterministic method, probabilistic analyses of the cases in the database are conducted using the Bayesian mapping function approach. The results of the probabilistic analyses, expressed as a mapping function, provide a simple means for probability-based evaluation of the liquefaction potential. The newly developed simplified method compares favorably to a widely used existing method.
|Number of pages||15|
|Journal||Journal of Geotechnical and Geoenvironmental Engineering|
|Publication status||Published - 2003 Jan 1|
All Science Journal Classification (ASJC) codes
- Environmental Science(all)
- Geotechnical Engineering and Engineering Geology