Scholarlink

Purpose <p>The engineering design of modern bridges is generally challenging task due to various performance requirements such as static and dynamic stability, reliability, robustness, and durability. Additionally, analysis of complex bridge structures is handled by using different vehicle-bridge interaction (VBI) models by acknowledging parameter uncertainties. This paper aims to evaluate aleatoric and epistemic uncertainties in the dynamic responses, which remain as an unattended work so far.</p> Methods <p>In this study, the stochastic dynamic analysis of bridge structures includes vehicle-bridge interaction as well as data- and model-related uncertainties. The latter is realized by using the generalized polynomial chaos (gPC) expansion method. A simply supported bridge structure serves as the basic structural model, where the finite element method solves the complex problem of quantifying the effects of moving vehicles, pothole excitations, and aleatoric (data-related) and epistemic (model-related) uncertainties. Different VBI models, e.g., the moving load model, moving mass model, and sprung mass model, are used to derive and analyze the response of a bridge system.</p> Results and Conclusion <p>The key outcomes are the time-dependent midspan displacement and midspan acceleration of the bridge and probability density functions (PDFs) of these responses with different levels of uncertainty (i.e., σ = 0.1 and σ = 0.2). The results indicate that the input uncertainty level and the variability of the VBI models significantly influence epistemic uncertainty, which is the least when utilizing a normal truncation in the gPC expansion model. Last but not least, the outcomes demonstrate that the proposed approach can be used by responsible structural engineering teams to analyze a variability of model uncertainties in an early development stage, resulting in a more efficient design process.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Stochastic Dynamic Response of Vehicle-Bridge Interaction Systems: Aleatoric and Epistemic Uncertainty Quantification

  • 期刊论文

  • 发表日期：

    2025年4月2日

查看全文
  • Manish Kumar Sah,
  • Sourav Chandra,
  • Marcus Maeder,
  • Vasant Matsagar,
  • Alessandro Palmeri,
  • Steffen Marburg

摘要

Purpose

The engineering design of modern bridges is generally challenging task due to various performance requirements such as static and dynamic stability, reliability, robustness, and durability. Additionally, analysis of complex bridge structures is handled by using different vehicle-bridge interaction (VBI) models by acknowledging parameter uncertainties. This paper aims to evaluate aleatoric and epistemic uncertainties in the dynamic responses, which remain as an unattended work so far.

Methods

In this study, the stochastic dynamic analysis of bridge structures includes vehicle-bridge interaction as well as data- and model-related uncertainties. The latter is realized by using the generalized polynomial chaos (gPC) expansion method. A simply supported bridge structure serves as the basic structural model, where the finite element method solves the complex problem of quantifying the effects of moving vehicles, pothole excitations, and aleatoric (data-related) and epistemic (model-related) uncertainties. Different VBI models, e.g., the moving load model, moving mass model, and sprung mass model, are used to derive and analyze the response of a bridge system.

Results and Conclusion

The key outcomes are the time-dependent midspan displacement and midspan acceleration of the bridge and probability density functions (PDFs) of these responses with different levels of uncertainty (i.e., σ = 0.1 and σ = 0.2). The results indicate that the input uncertainty level and the variability of the VBI models significantly influence epistemic uncertainty, which is the least when utilizing a normal truncation in the gPC expansion model. Last but not least, the outcomes demonstrate that the proposed approach can be used by responsible structural engineering teams to analyze a variability of model uncertainties in an early development stage, resulting in a more efficient design process.