Innovative Frameworks for the Probabilistic Performance-Based Design of Inelastic Wind Excited Structures

Abstract

Current prescriptive design provisions are moving towards performance-based design (PBD) approaches in which system-level probabilistic measures are used to explicitly describe performance. One of the key challenges in applying PBD in wind engineering is the evaluation of structural behavior over a full range of hazard levels, including extreme windstorms that could cause inelastic response. Despite the abundance of methods for inelastic response characterization of seismically excited systems, application of these methods to wind engineering is computationally challenging due to the extremely long duration of windstorms for which complex failure mechanisms can occur. The need to propagate uncertainties through the system in order to estimate the performance metrics further complicates the problem.

To address this situation, this research presents a performance-based wind engineering framework that integrates system-level collapse susceptibility models with non-collapse performance assessment frameworks. In particular, an efficient approach is proposed to rapidly estimate the inelastic behavior of wind excited systems based on the theory of dynamic shakedown. Safety against collapse susceptibility is then defined as the structure reaching the state of dynamic shakedown therefore ensuring safety against failure mechanisms such as low-cycle fatigue and/or incremental plastic collapse. To further account for failure due to excessive plastic deformations, a path-following algorithm is introduced for direct estimation of the residual displacements and plastic strains occurring at shakedown. The efficiency in estimating inelastic responses for any given wind load trace further allows simulation models to be directly applied for propagating uncertainty through the system and consequently estimating not only system-level reliability but also any other system-level performance metric of interest. A full scale archetype building is then studied with the aim of understanding, as compared to conventional elastic design procedures, the possibility of designing wind excited buildings to have controlled plasticity at the ultimate load level.

Finally, in addition to this mechanics-based approach, alternative data-driven approaches are developed through applying advanced metamodeling techniques for estimating the dynamic responses of uncertain wind excited systems. The potential of this data-driven approach is investigated on a high-dimensional building system subject to stochastic wind excitation and system uncertainty.