No abstract

This paper proposes a framework for establishing quantitative measures and mathematically reproducible definitions of structural resiliency as it pertains to a structure's ability to minimize the potential for undesirable response to low-probability-high-consequence events. The resiliency assessment and design process follow a logical progression of steps starting with the characterization of hazards and continuing through analysis simulations, damage modeling, and loss assessment by balancing functional relationships between design tradeoffs and associated consequences. The outcomes of each subprocess are articulated through a series of generalized variables: topology, geometry, damage, and hazard intensity measures. A rigorous probabilistic framework permits consistent characterization of the inherent uncertainties throughout the process. The proposed framework is well suited to support the building design process through stochastic characterization of assessment measures. Using a stepwise approach, the framework facilitates a systemwide method to confront multihazard threat scenarios by establishing functional relationships between the development of appropriate models, design methods, damage acceptance criteria, and tools necessary for implementation. The proposed methodology can be implemented directly for assessment of project-specific performance criteria or can be used as a basis for establishing appropriate performance criteria and provisions to achieve resilient structural solutions at the outset of design.

The possibility of local structural failure to propagate into global collapse of the structural system has fueled the continued development of improved computational methods to model building behavior, as well as "best practices" engineering standards. In spite of these efforts, recent events are bringing the issue of collapse mitigation to the forefront and highlighting the shortcomings of existing design practices. The catastrophic nature of structural collapse dictates the need for more reliable methodologies to quantify the likelihood of structural failures and strategies to minimize potential consequences.This paper presents the results of a stochastic nonlinear dynamic analysis study of a simple structural model to predict catastrophic failure. The performed analysis indicates that, at the point of incipient failure, uncertainties associated with the analysis become increasingly large. Consequently, it is not be possible to accurately predict when (and if) failure may occur. Recognizing the need to understand uncertainties associated with risk and occurrence of lowprobability-high-consequence events, this paper sets the stage to better understand the limitations of current numerical analysis methods and discuss innovative alternatives.

No abstract

No abstract