This paper aims to present a reliability analysis of the slab panel method (SPM) for the design of composite steel floors in severe fires. Rather than seeking to accurately define failure levels, this paper highlights areas of uncertainty in design and their effect on design results, whilst providing approximate reliability levels.
A Monte Carlo simulation has been conducted using the SPM design procedure to produce probability density functions of floor capacity for various floor layouts. Statistical input variables were obtained from the literature. Different configurations, geometries and fire severities are included to demonstrate how predicted floor capacities are influenced.
From the research presented, it is clear that the predicted reliability of SPM systems varies relative to a large number of criteria, but especially parameters related to fire loading. Predicted capacities are shown to be conservative compared to results of furnace and large-scale natural fire tests, which exhibit higher fire resistance. Due to distinct fire hazard categories with associated input values, there are step discontinuities in capacity graphs.
Limited research has been done to date on the reliability of structures in fire as discussed in this paper. It is important to verify the reliability levels of systems to ensure that partial and global factors of safety are adequate. Monte Carlo simulations are shown to be effective for calculating the average floor capacities and associated standard deviations. The presentation of probability density functions for composite floors in severe fires is novel.
