The overall objective of the project is to explore and advance the state-of-the-art in performance-based engineering in the context of multi-hazard analysis for earthquakes and windstorms. Performance-based engineering is a maturing concept requiring structures to meet specified objectives depending on the intensity of earthquake/windstorm loads. Practical implementation of this concept is not possible without the availability of efficient methods for quantifying the uncertainty in structural loads and propagating this uncertainty into response statistics. The last two decades have seen significant developments in both the quantification of load uncertainty and its propagation through structural systems. However, research is needed to describe earthquake/wind loads probabilistically, quantify material/structural uncertainties, and calculate response statistics. The proposed research will enhance preliminary work done by the Principal Investigator under a previous NSF grant.
The project will address uncertainty quantification and propagation and integrate them in a practical methodology for multi-hazard analysis. Task 1 will develop novel probabilistic models for seismic/wind hazards and loads that are based on all available information and are capable of capturing with increased accuracy the essential features of these loads. Hazards and loads will be characterized probabilistically by two parameters: moment magnitude and source-to-site distance for seismic loads and speed and direction for wind loads. Task 2 will develop conceptually simple, accurate, non-intrusive, and computationally efficient methods for calculating response statistics, fragility and damage/cost/downtime probability distributions, for structural systems in the linear/nonlinear regime under Gaussian/non-Gaussian loads. The method is based on stochastic reduced order models, and can be viewed as a smart Monte Carlo simulation. Task 3 will use results from previous two tasks to implement a practical methodology for performance-based multi-hazard engineering for seismic and wind loads. The task will calculate probability distributions for damage, cost, downtime, robustness, recovery, and other performance/resilience metrics. Robustness and recovery characterize the ability of structural/non-structural systems to withstand extreme events with limited loss of function and the time needed to restore the functionality of these systems to their pre-event level. Probability distributions, rather than expectations, will be used since they provide refined tools for decision making.
