LCE: Reliability-Based Simulation for Performance-Based Design

Johns Hopkins University, Baltimore MD

Investigators

Abstract

This Engineering Sciences for Modeling and Simulation-Based Life-Cycle Engineering (LCE) grant provides funding for the development of advanced computational tools to forecast the response of complex structural systems to extreme natural phenomena hazards, to analyze uncertainties, and to manage risk. Time-dependent structural reliability methods will be developed to provide the interfaces between models of the uncertain environment, physical models of the structural system, and its performance requirements. Databases and other information from experiment or field data, model simulation, and expert opinion will be incorporated using a computational Bayes framework. Highly efficient techniques for evaluating structural performance under uncertainty using sharpened methods for determining inclusion-exclusion bounds, will be investigated. The methods will be demonstrated using a testbed problem of significant current interest: fragility modeling of building structural systems subjected to severe earthquake stochastic ground motions. If successful, the decision tools developed in this research will provide a living picture of a complex civil infrastructure facility during its service life of 50 to 100 years, which could be used as a basis for the new performance-based design paradigm and for managing facility risk due to extreme natural hazards. In recent post-disaster assessments of building performance during earthquakes and hurricanes, it has been found that current design practices do not prevent non-life-threatening forms of structural damage and functional disruptions that may lead to direct economic and opportunity costs that approach or exceed the cost of the structure. The public is demanding engineering solutions with more predictable and controlled performance, greater durability, and lower life-cycle costs. A demonstration of advanced decision-theoretic technologies on an appropriate civil infrastructure testbed problem would accelerate improvements and enhance competition in building technologies and in structural design, minimize the need for prototype testing, and lower the risk of death, injury and property damage from natural hazards.