- Main
Application of Liquefaction Susceptibility Criteria within a Logic Tree Framework
- Sahin, Arda;
- Jana, Amalesh;
- Ulmer, Kristin J;
- Brandenberg, Scott J;
- Evans, T Matthew;
- Kramer, Steven L;
- Stewart, Jonathan P;
- Stuedlein, Armin W
Published Web Location
https://doi.org/10.1061/9780784486504.005Abstract
The probabilistic assessment of seismic ground failure hazards at any given site typically follows a three-step process that includes (1) estimating the probability of a given soil unit being susceptible to liquefaction (with the complement representing the probability of strength loss associated with cyclic softening); (2) estimating the probability that liquefaction would be triggered (if susceptible) or that cyclic softening would occur (if non-susceptible) given soil characteristics and anticipated ground shaking levels; and (3) estimating the potential impacts (surface manifestation, deformations, instabilities) from liquefied or softened strata. The three-stage liquefaction and cyclic softening hazard assessment process can be illustrated using a logic tree to allow different models and model weights to be considered in each step. We propose a logic tree formulation in which different weights are used for cyclic strength evaluation procedures for susceptible and non-susceptible soils. We posit that the procedures for evaluating the cyclic strength of sand-like soils should be different from the procedures used to evaluate the cyclic strength of clay-like soils, and that the weights assigned to these procedures should therefore be different. This approach adjusts the cyclic strength evaluation procedures—and their corresponding weights—based on the susceptibility category at the preceding node and the probability of susceptibility. A case history from the Next Generation Liquefaction database is used to illustrate the outcome of the proposed logic tree approach, with a focus on the influence of susceptibility on cyclic strength. The results demonstrate that the evaluation of susceptibility strongly influences the ultimate outcome of the analyses.
Many UC-authored scholarly publications are freely available on this site because of the UC's open access policies. Let us know how this access is important for you.