Uncertainty quantification and probabilistic performance-based assessment of nonlinear behavior of moment resisting frames under design and extreme wind loads

Abstract

Recent advancements in performance-based assessment and design of buildings subjected to wind loads call for the incorporation of uncertainties in the simulation of both loading and structural response. Additionally, they require a better understanding of nonlinear wind-induced response of buildings. Probabilistic analysis approaches are commonly used for the incorporation of uncertainties in performance-based assessment of building response. These approaches, however, focus more on the wind loading side as opposed to incorporating detailed nonlinear models for the structures. This study leverages SimCenter's newly developed tool WE-UQ to integrate a detailed nonlinear structural model of a 20-story 2D building frame into larger probabilistic simulations. The detailed structural model is created in finite element software OpenSEES and can simulate complex behavior such as stiffness and strength degradation and provide reasonably accurate predictions of yield and collapse. Simulation results are then utilized for probabilistic performance-based assessment of the building frame using the new guidelines proposed by the ASCE pre-standard for performance-based wind design. Additionally, performance of the frame under higher wind speeds that would lead to yielding or collapse of the building is also assessed. The results obtained from the probabilistic simulations are compared with pure deterministic simulation outcomes to investigate the effects and necessity of adding uncertainties into the response prediction of buildings subjected to wind loads.