Answers are grounded in this paper and cite their sources.
Structural reliability assessments have primarily focused on new structures. However, many existing structures may not meet current design codes, have been in service for decades, and have shown satisfactory performance after intense dynamic events, yet their reliability based on observed performance has received little attention. This study presents several key contributions: (1) it proposes three probabilistic approaches for reliability assessment of existing nonlinear structures, depending on the type and amount of information available. The first approach provides an approximate reliability estimate based solely on satisfactory performance after an intense dynamic event when available structural information is very limited. The second approach applies Bayesian updating to the probability distributions of mechanical properties by incorporating design data and field measurements; structural reliability is then assessed using these updated distributions. The third approach, representing the main contribution, performs direct Bayesian updating of the structural reliability model for existing nonlinear structures by incorporating actual observed performance after real dynamic events, addressing a critical gap in the literature. (2) It establishes a direct link between real observed performance and the intensity of dynamic loading within a theoretical safety margin framework. (3) It classifies the sources of uncertainty, enabling tailored mathematical analysis for each scenario. The proposed formulations are illustrated using two reinforced concrete buildings subjected to seismic disturbances. Collectively, these approaches provide a unified and rational framework for evaluating the reliability of existing structures subjected to real dynamic events, reducing implicit epistemic uncertainty and supporting more informed, probabilistic decision-making, thereby providing increased belief in structural reliability estimates.
No takes yet. Share an insight, caveat, or question.
Alamilla et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: