Computationally efficient stochastic approach for the fragility analysis of vertical structures subjected to thunderstorm downburst winds

Thunderstorm downburst winds introduce considerable uncertainty in dynamical structural analysis because of wind load non-stationarity, which cannot be adequately modelled by conventional stationary wind simulations. Performance-based structural analysis in wind engineering practice, which considers uncertainty related to both error propagation and modeling simplifications, requires sampling the structural information from a large number of dynamic simulations. This task may be computationally intensive using traditional numerical integration techniques. This study examines the feasibility and advantages of utilizing a wavelet-Galerkin (WG) approach to numerically integrate the coupled stochastic dynamic equations of motion for tall building structures affected by thunderstorm wind loads. The study examines the stochastic maximum structural response at key locations. Fragility analysis is subsequently conducted using curves modelled with log-normal complementary cumulative distribution functions and surfaces modelled using logistic regression. Both a “point-like” (plate) structure and a benchmark tall building are used for verification of the proposed simulation approach.