Probabilistic analysis of corrosion of reinforcement in RC bridges considering fuzziness and randomness

Lei Wang, Yafei Ma, Jianren Zhang, Yongming Liu

Research output: Contribution to journalArticlepeer-review

23 Scopus citations


A general methodology for probabilistic corrosion analysis of reinforcing bar in RC bridges is proposed in this paper. Uncertainties due to limited number of experimental data, incomplete inspection information, as well as the intrinsic randomness of random variables affect the prognostics of corrosion damage. The proposed study includes both fuzziness and randomness to consider different types of uncertainties. First, the chloride-induced corrosion initiation and propagation model are developed for reinforcement in RC bridge. The relationship between the area corrosion rate and the yield strength degradation is proposed on the basis of experimental investigation from accelerated corrosion testing. Following this, the randomness and fuzziness are included using fuzzy random variables to consider the uncertainties of the degradation under corrosive environments. The probabilistic modeling of the mean and the SD of reinforcement yield strength is discussed in detail. Finally, the proposed methodology is illustrated and validated with field measurements of corroded reinforcement in RC beams from a 34-year-old decommissioned RC bridge. The predicted results agree well with experimental values.

Original languageEnglish (US)
Pages (from-to)1529-1540
Number of pages12
JournalJournal of Structural Engineering (United States)
Issue number9
StatePublished - Sep 1 2013


  • Bridge
  • Corrosion
  • Fuzziness
  • Randomness
  • Reinforcing bar

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • General Materials Science
  • Mechanics of Materials
  • Mechanical Engineering


Dive into the research topics of 'Probabilistic analysis of corrosion of reinforcement in RC bridges considering fuzziness and randomness'. Together they form a unique fingerprint.

Cite this