This dissertation focuses on developing, evaluating and testing a new approach to seismic risk assessment of highway transportation networks. Throughout the years many researches have been conducted on each different step used in this dissertation. Most of these steps use a probabilistic approach for each single problem; Monte Carlo simulation is the brute force approach used by this dissertation, however some assumptions throughout the process were necessary to achieve results that could be useful to the final user, decision maker.;When dealing with the damageability of an infrastructure network, it is necessary to start from the study of the damageability of each vulnerable component. For highway transportation networks subject to a seismic event bridges are the vulnerable component. With enough time, enough information and enough manpower the best approach is step by step: to build an analytical model for each bridge, to perform probabilistic seismic hazard analysis PSHA at each bridge site and to assess the seismic vulnerability of each bridge.;Unfortunately, when dealing with large networks, bridges are present in hundreds or even thousands (LA-OC network has more than 3000 bridges; therefore subdivision in classes and prototyping of each class is necessary to reduce the problem to a more manageable dimension. a single prototype that represents its entire class is analyzed and its damageability is applied to the entire class. Fragility curves will be used to represents the bridges' damageability.;Different contributions are developed in this dissertation; the fundamental one is the development of fragility curves with two uni-variate random variables. The first will be the Intensity Measure of the ground motion; the second will be the angle of seismic incidence from which the ground motion struck the structure. Bridges with the same damageability class, located at different places within the network will behave differently under the same earthquake scenario, even if the intensity measure is the same, because of their different axis orientation and angle of seismic incidence. Therefore the spatial distribution properties of the network are taken into account when performing SRA.;The development of a new software was necessary for the application of these new fragility curves to the SRA of highway transportation networks.
展开▼
