Advanced search
1 file | 619.32 KB Add to list

Updating partial factors for material properties of existing structures in a Eurocode framework using Bayesian statistics

Robby Caspeele (UGent) and Luc Taerwe (UGent)
Author
Organization
Abstract
Bayesian statistics provide a powerful tool to update prior information based on additional information. In case of existing concrete structures, such a Bayesian approach is very attractive as often many uncertainties have to be dealt with in the structural assessment process. As material properties are very often considered as normal or lognormal variables, normal-gamma or lognormal-gamma distributions form a class of natural conjugate priors that enable to easily update prior hyperparameters for these properties. Further, these updated parameters can be used to update also partial factors which are used in a semi-probabilistic design/verification format as currently applied in the Eurocodes. A simplified Level II approach is used here to develop this methodology. As a result, a coherent semi-probabilistic evaluation procedure for existing structures is suggested, which consideres the current framework of the Eurocodes and enables to incorporate updated information based on e.g. on-site inspection data and data from testing, as these considerably influence the partial factors in the structural assessment process of existing structures. The suggested framework enables also to take into account different target safety levels, which can be based on ad-hoc cost optimization and the considered remaining working life. Finally, a short example is provided with respect to concrete compressive strength for which additional test results are available from compressive testing on concrete cores.

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 619.32 KB

Citation

Please use this url to cite or link to this publication:

MLA
Caspeele, Robby, and Luc Taerwe. “Updating Partial Factors for Material Properties of Existing Structures in a Eurocode Framework Using Bayesian Statistics.” Advances in Safety, Reliability and Risk Management : ESREL 2011, edited by Christophe Bérenguer et al., Taylor & Francis, 2011, pp. 1941–48.
APA
Caspeele, R., & Taerwe, L. (2011). Updating partial factors for material properties of existing structures in a Eurocode framework using Bayesian statistics. In C. Bérenguer, A. Grall, & C. Guedes Soares (Eds.), Advances in safety, reliability and risk management : ESREL 2011 (pp. 1941–1948). London, UK: Taylor & Francis.
Chicago author-date
Caspeele, Robby, and Luc Taerwe. 2011. “Updating Partial Factors for Material Properties of Existing Structures in a Eurocode Framework Using Bayesian Statistics.” In Advances in Safety, Reliability and Risk Management : ESREL 2011, edited by Christophe Bérenguer, Antoine Grall, and Carlos Guedes Soares, 1941–48. London, UK: Taylor & Francis.
Chicago author-date (all authors)
Caspeele, Robby, and Luc Taerwe. 2011. “Updating Partial Factors for Material Properties of Existing Structures in a Eurocode Framework Using Bayesian Statistics.” In Advances in Safety, Reliability and Risk Management : ESREL 2011, ed by. Christophe Bérenguer, Antoine Grall, and Carlos Guedes Soares, 1941–1948. London, UK: Taylor & Francis.
Vancouver
1.
Caspeele R, Taerwe L. Updating partial factors for material properties of existing structures in a Eurocode framework using Bayesian statistics. In: Bérenguer C, Grall A, Guedes Soares C, editors. Advances in safety, reliability and risk management : ESREL 2011. London, UK: Taylor & Francis; 2011. p. 1941–8.
IEEE
[1]
R. Caspeele and L. Taerwe, “Updating partial factors for material properties of existing structures in a Eurocode framework using Bayesian statistics,” in Advances in safety, reliability and risk management : ESREL 2011, Troyes, France, 2011, pp. 1941–1948.
@inproceedings{2131066,
  abstract     = {{Bayesian statistics provide a powerful tool to update prior information based on additional information. In case of existing concrete structures, such a Bayesian approach is very attractive as often many uncertainties have to be dealt with in the structural assessment process. As material properties are very often considered as normal or lognormal variables, normal-gamma or lognormal-gamma distributions form a class of natural conjugate priors that enable to easily update prior hyperparameters for these properties. Further, these updated parameters can be used to update also partial factors which are used in a semi-probabilistic design/verification format as currently applied in the Eurocodes. A simplified Level II approach is used here to develop this methodology. As a result, a coherent semi-probabilistic evaluation procedure for existing structures is suggested, which consideres the current framework of the Eurocodes and enables to incorporate updated information based on e.g. on-site inspection data and data from testing, as these considerably influence the partial factors in the structural assessment process of existing structures. The suggested framework enables also to take into account different target safety levels, which can be based on ad-hoc cost optimization and the considered remaining working life. Finally, a short example is provided with respect to concrete compressive strength for which additional test results are available from compressive testing on concrete cores.}},
  author       = {{Caspeele, Robby and Taerwe, Luc}},
  booktitle    = {{Advances in safety, reliability and risk management : ESREL 2011}},
  editor       = {{Bérenguer, Christophe and Grall, Antoine and Guedes Soares, Carlos}},
  isbn         = {{9780415683791}},
  language     = {{eng}},
  location     = {{Troyes, France}},
  pages        = {{1941--1948}},
  publisher    = {{Taylor & Francis}},
  title        = {{Updating partial factors for material properties of existing structures in a Eurocode framework using Bayesian statistics}},
  year         = {{2011}},
}