Stochastic Analysis to Predict Reliability Index of a Tall Building Structure

Résumé

In this paper, stochastic analyses are performed to predict the failure probability and reliability index of tall building structures with random parameters under random seismic loading conditions using traditional  Monte Carlo Simulation (MCS) method. Randomness of geometrical and structural properties of materials as well as seismic loading uncertainties are considered. The sensitivity of structural reliability is examined in light of various values of the limit level of the performance variable. Numerical results show that structural reliability is generally affected by the variability of all uncertain parameters but more importantly by seismic loading randomness. The effects on structural reliability are shown to be more pronounced for higher variability of the stochastic variables. In addition, preliminary sensitivity analysis based on the First Order Reliability Method (FORM) that gives information on the sensitivity of the randomness of the inputs parameters, shows that the original 11 random variable seismic reliability problem can be effectively reduced utilizing only 4 random variables, namely: the Peak Ground Acceleration, concrete elastic modulus, core inertia and reinforced concrete density.

Mots-clés

Auteurs

1. 1Departement of civil Engineering, Ecole nationale polytechnique