Click the title below to display the complete page!
Evaluation of liquefaction susceptibility is affected by different sources of uncertainties. These uncertainties may be due to the applied models as well as the limited availability of data from soundings and differences in quality. In the context of this paper, the uncertainties will be discussed based on both a deterministic and a probabilistic approach, using state-of-the-art methods for evaluation of liquefaction susceptibility in combination and stand alone. Firstly, deterministic procedures using laboratory criteria and the simplified procedure are discussed based on their sensitivity to the input parameter. Second, the sensitivity of the input parameter is studied in a probabilistic analysis, while uncertainties of the soil parameters have been incorporated into a Bayesian Probabilistic Network. Third, the results of the deterministic analysis are compared with the probabilistic analysis based on available data from the city of Adapazari in Turkey, where examples of damage caused by liquefaction were numerous. Both approaches are difficult to compare due to controversial results and the different procedures used. Additionally, the deterministic methods adopted might not be consistent in their terminology for initiation of liquefaction using the terms “susceptible to liquefaction”, “liquefaction possible” and “liquefaction very likely”. Conclusions for further activities in this field will be drawn.
Authors:Buchheister, Juliane and Bayraktarli, Yahya and Laue, Jan and Faber, Michael
Index Terms:SoilGroup; Merci; uncertainty; liquefaction; Bayraktarli, Yahya; Buchheister, Juliane; Faber, Michael; Laue, Jan
Further Information:Date published: 03.09.2006