A Probabilistic Safety Evaluation Framework for Multi-Hazard Assessment in a Bridge using SO-MARS Learning Model

A probabilistic evaluation procedure is established to assess a bridge safety against floods and earthquakes, which are the two major threats of a bridge in Taiwan. Scour depth distribution is used to reflect the flood hazard, in which swarm optimized multivariate adaptive regression splines (SO-MARS) is utilized to calculate scour depth density followed by Monte Carlo Simulation (MCS) and a scour risk curve is constructed. Displacement ductility is used to measure the bridge performance under attacks of both hazards through nonlinear time history analyses followed by a power law regression to build the fragility curve. A code-based probabilistic seismic hazard curve is constructed and the joint failure probability under seismic and flood attacks is obtained. A numerical example is provided to illustrate the proposed methodology, in which the nonlinear behaviors in concrete (including core and cover areas), steel bar and soil are included in a bridge model. A threshold scour depth for a given earthquake (e.g., the return period or the peak ground acceleration is given) is derived to meet a specified target reliability. The suggested scour depth is a deterministic number which is immediately applicable in engineering practice.