Λ-factors updating for seismic analysis and design of base-isolated bridges

Seismic isolation has been used for decades to protect new and existing bridges from earthquakes. Nevertheless, the effect of the mechanical variability of high damping rubber bearings on the reliability of these strategic structures is still a research topic, especially in the light of the severe environmental condition and working life that must be considered in bridge design. Most worldwide advanced seismic codes prescribe upper bound and lower bound analyses to account for the potential variations of the bearing response, however, common values currently used are still based on outdated test campaigns. Moreover, the actual effect of these variabilities on the global reliability of the isolated structure is not well assessed yet. To investigate this topic, a comparative probabilistic analysis is carried out to evaluate the influence of different sets of modification factors on the probability of failure. More in detail, modification factors currently provided by EN15129 and a new modification factors derived from a set of recent experimental data is considered, and the comparison is carried out by inspecting the demand hazard curves computed through a full probabilistic analysis accounting for both the bearing properties variability and the seismic input variability. Subset Simulation method in combination with a stochastic model for the bidirectional horizontal ground motion has been employed to achieve an accurate estimates of probability of failure. To reduce the computational effort of this full probabilistic approach, a 3D-model with a reduced number of degrees of freedoms is adopted. Results highlight the effect of variability in the full range of significant rate of exceedance of earthquake intensities, offering preliminary insights on the actual effects of mechanical variability on the reliability of this kind of structures.