This paper proposes a framework for mixed uncertainty quantification in the remaining service life assessment of existing deteriorated reinforced concrete (RC) bridges under temperature climate changes. The probability bound analysis (PBA) is deployed for quantifying mixed uncertainty. The epistemic uncertainty in climate change is modelled using a distribution-free p-box in the PBA framework. An extension of Chebyshev’s inequality is proposed to construct a parametric distribution-free probability-box (p-box) under a specific confidence interval. This ensures consistency with the distributional p-box, constructed based on the uncertainty in distribution parameters. The proposed method can account for parametric uncertainties for variables with known or unknown distribution types with a specified confidence level and provides narrower bounds than other approaches. The types and sources of uncertainty associated with assessing an existing bridge are thoroughly investigated as an application example. An insight into how epistemic uncertainty governs corrosion initiation and propagation in service life prediction is provided. Results show that the epistemic uncertainty in the diffusion coefficient is most critical in the probabilities of corrosion initiation and damage, followed by the epistemic uncertainty in temperature projections.

中文翻译：

---

考虑温度气候变化模型中的不确定性，对暴露于氯化物诱导腐蚀的老化基础设施进行安全评估


本文提出了一种在温度气候变化下现有劣化钢筋混凝土 （RC） 桥梁剩余使用寿命评估中的混合不确定性量化框架。概率界分析 （PBA） 用于量化混合不确定性。气候变化的认识不确定性是使用 PBA 框架中的无分布 p-box 建模的。提出了切比雪夫不等式的扩展，以在特定置信区间下构建参数化无分布概率框 （p-box）。这确保了与基于分布参数的不确定性构建的分布 p-box 的一致性。所提出的方法可以解释具有指定置信度的已知或未知分布类型的变量的参数不确定性，并提供比其他方法更窄的边界。作为应用示例，对与评估现有桥梁相关的不确定性的类型和来源进行了彻底调查。本文深入探讨了认识不确定性如何控制使用寿命预测中的腐蚀萌生和传播。结果表明，扩散系数的认识不确定性在腐蚀开始和破坏的概率中最为关键，其次是温度预测的认识不确定性。