Estimating the Severity of Liquefaction Ejecta Using the Cone Penetration Test

Hutabarat, Daniel; Bray, Jonathan D.

doi:10.1061/(asce)gt.1943-5606.0002744

Cited by 17 publications

(14 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Simultaneously, this work demonstrated the insufficiency of the H 1 -H 2 model variables and the role that mechanical knowledge must play in generating better predictors. To that end, the tier-2 manifestation model of Hutabarat and Bray (2022) uses novel and insightful predictors inspired by tier-3 modeling results, but ultimately, is resigned to conventional empiricism when developing and calibrating a model. This is another example of how AI might improve the empirical component(s) of a model while also respecting and benefiting from human knowledge of mechanics.…”

Section: Applications and Models That Should Not Be Ignoredmentioning

confidence: 99%

“…Because triggering refers to liquefaction at discrete depths in a profile, the outputs from triggering analysis are often used in series with manifestation models to predict a profile’s system response in the form of settlement, ejecta, spreading, cracking, and so on (e.g. Hutabarat and Bray, 2022; Iwasaki, 1978; Maurer et al, 2015; Youd et al, 2002; Zhang et al, 2002). The third tier, “tier 3,” is defined by “constitutive” models that predict liquefaction response at the soil element level, typically using numerous input parameters.…”

Section: Literature Overview and Backgroundmentioning

confidence: 99%

“…Although efforts to understand manifestation mechanics have increased in the wake of the Canterbury sequence and should continue (e.g. Bassal and Boulanger, 2021; Cubrinovski et al, 2019; Hutabarat and Bray, 2022), these efforts have either not yet produced new tier-2 models or the resulting models have not yet been shown to outperform existing methods. These efforts have also highlighted the potential advantages of tier-3 models for producing insights into manifestation patterns and severities, and thus, may instead benefit the profession by increasing the use of more advanced methods.…”

Section: Applications and Models That Should Not Be Ignoredmentioning

confidence: 99%

See 2 more Smart Citations

Why “AI” models for predicting soil liquefaction have been ignored, plus some that shouldn’t be

Maurer

Sanger

2023

Earthquake Spectra

View full text Add to dashboard Cite

Soil liquefaction remains an important and interesting problem that has attracted the development of enumerable prediction models. Increasingly, these models are utilizing algorithmic learning or “artificial intelligence” (AI). The rapid growth of AI in the liquefaction literature is unsurprising, given its ease of implementation and potential advantages over traditional statistical methods. However, AI liquefaction models have been widely ignored by practitioners and researchers alike; the objective of this article is to investigate “why?” Through a sample review of 75 publications, we identify several good reasons. Namely, these models frequently (1) are not compared to state-of-practice models, making it unclear why they should be adopted; (2) depart from best practices in model development; (3) use AI in ways that may not be useful; (4) are presented in ways that overstate their complexity and make them unapproachable; and (5) are discussed but not actually provided, meaning that no one can use the models even if they wanted to. These prevailing problems must be understood, identified, and remedied, but this does not mean that AI itself is problematic or that all prior efforts have been without merit or utility. Instead, understanding these recurrent shortcomings can help improve the direction and perceptions of this growing body of work. Toward this end, we highlight papers that are generally free from these shortcomings, and which demonstrate applications where AI is more likely to provide value in the near-term: permitting new modeling approaches and potentially improving predictions of liquefaction phenomena.

show abstract

Section: Applications and Models That Should Not Be Ignoredmentioning

confidence: 99%

Section: Literature Overview and Backgroundmentioning

confidence: 99%

Section: Applications and Models That Should Not Be Ignoredmentioning

confidence: 99%

See 1 more Smart Citation

Why “AI” models for predicting soil liquefaction have been ignored, plus some that shouldn’t be

Maurer

Sanger

2023

Earthquake Spectra

View full text Add to dashboard Cite

show abstract

“…This method only considers soils in the top 15 m of a soil profile and was developed using FS liq estimated from the Boulanger and Idriss (2016) triggering procedure at a probability level of 0.5. Hutabarat and Bray (2022) define threshold levels of L D and C R combinations that differentiate between liquefaction ejecta severity levels.…”

Section: Existing Manifestation Modelsmentioning

confidence: 99%

“…What this result shows then, is that when using the same input data and when trained on the same database of case histories, CBR can generate models with better predictive power than existing models. The Hutabarat and Bray (2022) model has a low R TP value because it was developed to estimate only manifestation due to ejecta, and not other forms of manifestation, such as cracking or settlement, which are included in the Global database. Therefore, even if it correctly predicts cases with ejecta manifestation, it misses other forms of manifestation because it was not strictly developed to predict their occurrence.…”

Section: Comparison With Existing Modelsmentioning

confidence: 99%

Evaluation of case-based reasoning to estimate liquefaction manifestation

Carlton,

Geyin,

Engin

2023

Earthquake Spectra

View full text Add to dashboard Cite

This article develops a framework for and explores the use of case-based reasoning (CBR) to predict seismically induced liquefaction manifestation. CBR is an artificial intelligence process that solves new problems using the known answers to similar past problems. CBR sorts a database of case histories based on their similarity to a design case and predicts the outcome of the design case as the observed outcome of the most similar case history or majority outcome of the most similar case histories. Two databases of liquefaction case histories are used to develop and validate numerous CBR models. Different input parameters and aspects of the CBR method and their influence on the predictive capability of the models are evaluated. Some of the developed CBR models were shown to have a better predictive power than currently existing models. However, more research is needed to refine these models before they can be used in practice. Nevertheless, this study shows the potential of CBR as a method to estimate liquefaction manifestation and suggests several avenues of future research.

show abstract