Durability reliability analysis for corroding concrete structures under uncertainty

• Published in: Mechanical systems and signal processing Vol. 101; pp. 26 - 37
• Main Author: Zhang, Hao
• Format: Journal Article
• Language: English
• Published: Berlin Elsevier Ltd 15.02.2018; Elsevier BV
• Subjects: Chloride; Concrete structures; Corrosion; Durability; Epistemic uncertainty; Imprecise probability; Interval analysis; Materials durability; Mathematical models; Probabilistic methods; Probability; Probability box; Random variables; Reinforced concrete; Reliability; Reliability analysis; Statistical analysis; Studies; Uncertainty; Interval analysis; Epistemic uncertainty; Uncertainty; Imprecise probability; Concrete structures; Corrosion; Probability box; Durability; Reliability
• ISSN: 0888-3270; 1096-1216
• DOI: 10.1016/j.ymssp.2017.08.027

•Epistemic uncertainties play an important role in the durability reliability of deteriorating reinforced concrete structures.•Ignoring the effects of epistemic uncertainty may lead to a false impression of the reliability.•Epistemic uncertainties can be modelled either by the imprecise probability approach or the Bayesian approach.•The durability reliability is most sensitive to the epistemic uncertainty in the chloride diffusion coefficient. This paper presents a durability reliability analysis of reinforced concrete structures subject to the action of marine chloride. The focus is to provide insight into the role of epistemic uncertainties on durability reliability. The corrosion model involves a number of variables whose probabilistic characteristics cannot be fully determined due to the limited availability of supporting data. All sources of uncertainty, both aleatory and epistemic, should be included in the reliability analysis. Two methods are available to formulate the epistemic uncertainty: the imprecise probability-based method and the purely probabilistic method in which the epistemic uncertainties are modeled as random variables. The paper illustrates how the epistemic uncertainties are modeled and propagated in the two methods, and shows how epistemic uncertainties govern the durability reliability.