• Document: Validation and Application of Empirical Liquefaction Models
  • Size: 1.15 MB
  • Uploaded: 2019-04-16 12:53:59
  • Status: Successfully converted


Some snippets from your converted document:

Validation and Application of Empirical Liquefaction Models Thomas Oommen1; Laurie G. Baise, M.ASCE2; and Richard Vogel, M.ASCE3 Abstract: Empirical liquefaction models 共ELMs兲 are the standard approach for predicting the occurrence of soil liquefaction. These models are typically based on in situ index tests, such as the standard penetration test 共SPT兲 and cone penetration test 共CPT兲, and are broadly classified as deterministic and probabilistic models. No objective and quantitative comparison of these models have been published. Similarly, no rigorous procedure has been published for choosing the threshold required for probabilistic models. This paper provides 共1兲 a quantitative comparison of the predictive performance of ELMs; 共2兲 a reproducible method for choosing the threshold that is needed to apply the probabilistic ELMs; and 共3兲 an alternative deterministic and probabilistic ELM based on the machine learning algorithm, known as support vector machine 共SVM兲. Deterministic and probabilistic ELMs have been developed for SPT and CPT data. For deterministic ELMs, we compare the “simplified procedure,” the Bayesian updating method, and the SVM models for both SPT and CPT data. For probabilistic ELMs, we compare the Bayesian updating method with the SVM models. We compare these different approaches within a quantitative validation framework. This framework includes validation metrics developed within the statistics and artificial intelligence fields that are not common in the geotechnical literature. We incorporate estimated costs associated with risk as well as with risk mitigation. We conclude that 共1兲 the best performing ELM depends on the associated costs; 共2兲 the unique costs associated with an individual project directly determine the optimal threshold for the probabilistic ELMs; and 共3兲 the more recent ELMs only marginally improve prediction accuracy; thus, efforts should focus on improving data collection. DOI: 10.1061/共ASCE兲GT.1943-5606.0000395 CE Database subject headings: Bayesian analysis; Validation; Empirical equations; Soil liquefaction; Data collection. Author keywords: Bayesian updating method; Support vector machine; Simplified procedure; Model validation; CPT; SPT; Machine learning. Introduction Toprak et al. 1999; Juang et al. 2002; Goh 2002; Cetin et al. 2002, 2004; Lee et al. 2003; Sonmez 2003; Lai et al. 2004; Sonmez and Soil liquefaction is the loss of shear strength induced by shaking, Gokceoglu 2005; Papathanassiou et al. 2005; Holzer et al. 2006; which can lead to various types of ground failures. Liquefaction is Moss et al. 2006; Juang and Li 2007兲. This paper attempts to most often evaluated with empirical liquefaction models 共ELMs兲. improve liquefaction models by 共1兲 quantitatively comparing the ELMs have been developed for in situ index tests, such as stan- predictive performance of several ELMs; 共2兲 identifying the dard penetration test 共SPT兲, cone penetration test 共CPT兲, and threshold needed to apply the probabilistic ELMs; and 共3兲 devel- shear-wave velocity 共Vs兲. These in situ data are used to estimate oping an alternative deterministic and probabilistic ELM based on the potential for “triggering” or initiation of seismically induced the machine learning algorithm, known as support vector machine liquefaction. In the context of the analyses of in situ data, the 共SVM兲. estimate of liquefaction potential derived from ELMs can be Currently, the most widely used ELM for the assessment of broadly classified as 共1兲 deterministic 共Seed and Idriss 1971; liquefaction potential is the “simplified procedure,” originally rec- Iwasaki et al. 1978; Seed et al. 1983; Robertson and Campanella ommended by Seed and Idriss 共1971兲 based on SPT blow counts. 1985; Seed and De Alba 1986; Shibata and Teparaksa 1988; Goh Since 1971, this procedure has been revised and developed for 1994; Stark and Olson 1995; Robertson and Wride 1998; Juang et other in situ tests, such as the CPT and Vs 共Seed et al. 1985; Youd al. 2000, 2003; Idriss and Boulanger 2006; Pal 2006; Hanna et al. and Noble 1997兲. Simplified methods that follow the general for- 2007; Goh and Goh 2007兲 and 共2兲 probabilistic 共Liao et al. 1988; mat of the Seed-Idriss procedure were reviewed in a workshop report edited by Youd and Noble 共1997兲. Youd et al. 共2001兲, Cetin 1 Postdoctoral Associate, Dept. of Civil and Environmental E

Recently converted files (publicly available):