Stochastic live load model for buildings and its application in reliability based code calibration

Resumo

While the exact loading to which a structure will be subjected cannot be precisely assessed during the
design phase due to its stochastic nature, probabilistic models are useful for the rational determination of nominal
values, partial safety factors and load combination factors employed in limit state design that accurately reflects the
variability of these loads. In this paper, a simple probabilistic model describing the spatial and temporal variabilities
of live loads in buildings is presented. This model consists of a sum of a sustained load and an intermittent load
stochastic processes. Due to the lack of national data to back up the model, parameters are taken from the Joint
Committee on Structural Safety (JCSS), based on international surveys. Using this stochastic model, sample values
for live loads are generated for buildings with different occupancy types, and statistics for the fifty-year extreme and
arbitrary point-in-time distributions of live loads are derived using Monte Carlo simulations. These values are then
compared with those of Brazilian design codes ABNT NBR 6120:2019 (Design Load for Structures), and other
major international standards. The resulting statistics are also employed in a reliability-based calibration of the
partial safety factors presented in Brazilian design codes for steel (ABNT NBR 8800:2008) and concrete (ABNT
NBR 6118:2014) structures. It is shown that, with the resulting set of optimized partial safety factors, reliability
is made more uniform over different load ratios, with a smaller variation around a chosen target reliability value,
while attaining no significant economic impact when compared with the currently employed factors.