Seismic Risk Analysis of Existing Link Slab Bridges Using Novel Fragility Functions

Abstract

In this paper, a comprehensive probabilistic framework is proposed and adopted to perform seismic reliability and risk analysis of existing link slab (LS) bridges, representing a widely diffused structural typology within the infrastructural networks of many countries worldwide. Unlike classic risk analysis methods, innovative fragility functions are used in this work to retrieve more specific and detailed information on the possible failure modes, without limiting the analysis to the global failure conditions but also considering several intermediate damage scenarios (including one or more damage mechanisms), and providing insights on the numerosity of elements involved within a given damage scenario. Reliability analyses are performed on a set of LS bridges with different geometries (total lengths and pier heights) designed according to the Italian codes enforced in the 1970s. Accurate numerical models are developed in OpenSees and Multiple-Stripe nonlinear time–history analyses are carried out to build proper demand models, from which fragility functions are determined according to two limit states: damage onset and near-collapse. Mean annual rates of exceeding are thus estimated through the convolution between the hazard and the fragility. The results shed light on the main failure mechanisms characterizing this bridge typology, highlighting how different levels of risk (hence safety margins) can be associated with failure scenarios that differ in terms of elements/mechanisms involved and damage extension. Such a higher level of detail in the risk analysis may be useful to better quantify post-earthquake consequences (e.g., costs and losses) and define more tailored retrofit interventions. A comparison of the reliability levels associated with bridges of the same class with different geometries is finally presented.

Influence of Knowledge Uncertainties on the Safety Assessment of Existing Post-tensioned Concrete Bridges

“…Due to different sources of uncertainties 21, the basic variables 𝒙 are described by a vector of random variables (r.v.) 𝑿 , whose statistical properties are provided by the joint probability distribution function 𝑓 𝑋 (𝒙), so that the failure probability 𝑃 𝑓 , which is a probabilistic measure of the structural performance 22, can be evaluated by: 𝑃 𝑓 = ∫ 𝐼(𝒙)𝑓 𝑋 (𝒙)𝑑𝒙 (1) where the index function 𝐼(𝒙) in Equation ( 1) is defined as follows: 𝐼(𝒙) = 1 if 𝑔(𝒙) < 𝑔̅ , otherwise 𝐼(𝒙) = 0. The system response is satisfactory if 𝑃 𝑓 < 𝑃 𝑓 ̅ in which the target value 𝑃 𝑓 ̅ , is usually based on the evaluation of economic and social consequences due to failure.…”