Optimum selection of retrofit measures for R/C bridges using fragility curves

The key objective of this paper is to present a methodology for the selection of the appropriate retrofit scheme for bridges, based on the performance of different components (piers, bearings, abutments) with a view to upgrading the seismic performance of the system. The methodology is applied to a common bridge type, with monolithic pier-to-deck connections. Threshold values for global and local parameters are properly defined based on performance criteria, for different damage states. The correlation of local and global threshold values calculated on the basis of member and system capacity, respectively, reveals the structural system's robustness and identifies the most critical member with respect to the seismic performance of the system. Alternative retrofit measures, i.e. reinforced concrete (R/C) and fibre-reinforced polymer (FRP) jackets, and bearing replacement, are applied to critical members of the structure. Fragility curves are generated for the resulting retrofitted bridges using a probabilistic seismic demand model based on the results of inelastic dynamic response history analysis for appropriately selected earthquake ground motions, while bridge capacity is estimated through pushover analysis. The engineering demand parameter used for the quantification of the limit state threshold values for the retrofitted bridge is the bridge displacement (global parameter), related to the most critical component's displacement for every limit state and to local parameters (such as the pier curvature ductility). Alternative sets of retrofit properties are considered, and material uncertainties are treated in a probabilistic way in the bridge model (Latin Hypercube Sampling Method). The alternative retrofit schemes are finally evaluated and the optimum retrofit solution is identified on the basis of its efficiency, emerging from the comparison of fragility curves (as-built and retrofitted case) for different levels of seismic action.