Mechanics & Industry
Volume 17, Number 6, 2016
|Number of page(s)||12|
|Published online||07 July 2016|
A comparative study of construction methods for seismic fragility curves using numerical simulations
1 Institut Pascal – Institut Français
de Mécanique Avancée, Campus de
2 Université Paris-Est, Laboratoire Navier (ENPC/IFSTTAR/CNRS), École des Ponts ParisTech, 6 & 8 av. Blaise Pascal, Champs-sur-Marne, 77455 Marne-la-Vallée Cedex 2, France
3 Danang University of Science and Technology, 54 Nguyen Luong Bang, Hoa Khanh, Lien Chieu, Danang, Vietnam
a Corresponding author:
Accepted: 16 December 2015
A seismic fragility curve of a structure or a mechanical system, presenting its failure probability versus seismic intensity, can be established by an engineering judgment approach, an empirical approach or a numerical approach. In the numerical approach, there exist three popular methods: the scaled seismic intensity method, the maximum likelihood estimation method, and the probabilistic seismic demand/capacity models. This paper is focused on a comparison of these numerical methods. Linear/non-linear oscillators and an eight-storey frame structure were used to derive fragility curves for different tests. The results obtained show a discrepancy of fragility curves between the methods, and their quality is commented in comparison with results of the Monte-Carlo method with a high number of simulations. The maximum likelihood estimation method is in general the recommended method, due to its good accuracy in both numerical examples.
Key words: Seismic fragility curve / numerical simulation / seismic intensity scaling / maximum likelihood estimation / probabilistic seismic demand and capacity models
© AFM, EDP Sciences 2016
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.