Performance-based assessment of steel girder bridges for open-air fire hazards

Abstract This dissertation proposes a decision-making approach to mitigate the effects of fire on steel girder highway bridges. The proposed assessment approach accounts for uncertainties pertaining to both the fire scenarios and bridge structures. It provides a performance envelope that indicates the risk of fire-induced damage, as well as the effectiveness of potential fire protection strategies. To enable this process, a computationally efficient method of simulating the performance of steel bridges subjected to open-air fire hazards is developed. The second generation modified discretized solid flame (MDSF2) fire model, which simulates the heat transfer from a hydrocarbon pool fire of varying size and fuel type to structural targets, is proposed and calibrated based on experimental data and high-fidelity computational simulations. As a demonstration, the fire hazard assessment procedure (which includes the fire model, heat transfer simulation, and structural assessment) is applied to a typical multi-span steel girder bridge to evaluate the fire resistance of the bridge superstructure (i.e. the girders, cross-frames, and deck slab). The fire-induced damage state is described using several performance indicators, such as maximum and residual deflection, local buckling, and lateral torsional buckling. Building on this demonstration, a fragility analysis is conducted for simple-span steel girder overpasses to provide probabilistic assessment of bridge damage due to fire hazards. Uncertainty associated with the fire location and combustion properties, structural properties of bridge, and gravity loading are considered via stochastic sampling from probabilistic distributions. Fragility functions are derived based on the results of these stochastic analyses for steel girder bridges with varying spans. The outcome of this dissertation can be used for fire hazard assessment and mitigation prioritization with limited resources according to flexible objectives.