Fire fragility functions are a powerful method to characterize the probabilistic vulnerability of buildings to fire in the context of urban resilience assessment. But this method is recent and the influence of the different uncertain parameters on the functions has not been systematically studied. The first objective of this paper is to identify the prevailing parameters in constructing fire fragility functions for steel frame buildings. To this end, sensitivity analyses are conducted using Monte Carlo Simulations and a variance-based method, focusing on column failure fragilities. Fragilities for buildings with 3 to 12 stories, 0 to 3 h fire resistance rating and various occupancies are compared, assuming compartment areas ranging from 15 m2 to 80 m2. Results show that uncertainties in fire, heat transfer and structural models all generate significant variability in the fire fragility. In addition to fire load as the intensity measure, significant probabilistic parameters are the compartment geometry and openings, the thickness and thermal conductivity of fire protection, and the temperature dependent mechanical properties of steel. The second objective is to clarify the incorporation of fragility functions in a comprehensive structural fire reliability framework. A methodology for combining the functions with the ignition likelihood per year and with the fire loading in MJ/m2 is described, yielding annual probability estimates of column failure due to fire in the buildings. For a sprinklered office building designed according to prescriptive provisions, this annual probability ranges from 1.90 × 10−7 to 0.12 × 10−7 per year as a function of the building height. The probabilistic modeling techniques proposed in this paper can be used to establish consistent reliability levels in different buildings and to evaluate resilience for fire scenarios.
All Science Journal Classification (ASJC) codes
- Materials Science(all)
- Safety, Risk, Reliability and Quality