IDESIGN · Enabling Seismic Design Decision-Making under Uncertainty

A robust methodology for performance-based seismic design is proposed that encompasses advanced analysis techniques under a common probabilistic framework to allow seismic design decision-making in the presence of uncertainties. Such a design method, whereby a structure is designed to satisfy a range of performance objectives paired with specific seismic hazard levels, has become the object of intensive research in the wake of the staggering economic losses witnessed in recent seismic events. Despite current advances in the adjacent field of performance-based assessment, and the emergence of displacement-based methods, no clear path has yet been defined on how to apply such results in design without considerable computational effort. At present, this necessitates a cumbersome process of analysis and redesign cycles that will slowly converge to a satisfactory, albeit largely non-optimal, structure. The task becomes more challenging when attempting to include the influence of epistemic uncertainties inherent in the structural model, analysis method and seismic loading.We will explore ways to offer a robust method for performance-based design using either static or dynamic nonlinear analysis techniques that can be readily used for practical applications and consequently implemented into seismic design code provisions. The fundamental building blocks will be (a) the formulation of standardized 2D and 3D mechanical models, (b) the quantification of the effect of uncertainties due to the modeling parameters and the analysis method used, (c) the use of a simplified closed-form probabilistic framework (d) the calibration versus optimal solutions and the estimation of the degree of conservatism involved in comparison to accurate assessment techniques. The ultimate goal is to develop a practical, yet accurate and safe, state-of-the-art performance-based design methodology, that will result to advantageous structural designs compared to the traditional force-based approaches.