Target reliability index optimization framework for multiple building classes based on community-level objectives. (July 2021)

• Record Type: Journal Article
• Authors: Gudipati, Vamshi Krishna; Cha, Eun Jeong
• Is Part Of: Structural safety. Volume 91(2021)
• Publication Date: 2021-07
• Subjects: Building class -- Community -- Interdependency -- Neural networks -- Optimal target reliability index -- Seismic; Structural stability -- Periodicals; Safety factor in engineering -- Periodicals; Reliability (Engineering) -- Periodicals; Constructions -- Stabilité -- Périodiques; Coefficient de sécurité en ingénierie -- Périodiques; Fiabilité -- Périodiques; 620.86
• DOI: 10.1016/j.strusafe.2021.102097
• Languages: English
• ISSNs: 0167-4730

Abstract

Highlights: Building target reliability index (TRI) optimization framework is introduced. The framework is developed for multi-building class TRI optimization. Community-level objectives are incorporated into the framework. Computational efficiency is achieved with neural network-based response estimation. Considering building loss interdependency is crucial in determining optimal TRI. Abstract: The impact of a disaster is spread over a region and thus the corresponding risks can be managed effectively by carefully considering the community impact. Since structural design codes regulate the safety levels of structures, it is appropriate to consider a community-level objective in the determination of the safety levels stipulated in the structural design codes. However, evaluating community disaster impact requires consideration of loss interdependency among different buildings, which could be a complex task. Furthermore, implementing a community-level objective to design safety level determination requires the establishment of feasible methods to calculate it in a reasonable time. This paper develops a simple and viable framework for community-level objective-based design safety target optimization for multiple building classes subjected to seismic hazard. To reduce the computation cost, the framework employs an efficient procedure of using an artificial neural network to approximate the seismic response of buildings with common characteristics (occupancy use, and type and size of lateral resisting system). The methodology is illustrated for optimizing the target reliability indices of office and hospital building classes exposed to seismic hazard, with minimizing regional economic cost as objective. Simple loss and interdependency models are used for this illustration due to lack of available complex models, which can be replaced by more recent and complex models to improve accuracy of the analysis once they become available.