Title: A probabilistic model for corrosion prediction of steel reinforcement

Authors: Wei-Liang Jin, Xiao-Zhou Wang, Zhi-Gang Song

Addresses: Department of Civil Engineering, Zhejiang University, Hangzhou 310027, P.R. China. ' Department of Civil Engineering, Zhejiang University, Hangzhou 310027, P.R. China. ' Department of Civil Engineering, Zhejiang University, Hangzhou 310027, P.R. China

Abstract: A path probability model (PPM) for predicting the probability distribution of steel reinforcement subject to chloride ingress is presented in this paper to evaluate the corrosion level in a probability way. The whole corrosion history of given service period is divided into a series of paths formed by initiation phase and propagation phrase, which are respectively associated to the probability of corrosion initiation and the conditional probability of steel corrosion ratio when the corrosion has initiated. Based on PPM, a dynamic simulation has been performed to predict the time-dependent probability density function of corrosion ratio using Monte Carlo simulation technique. The efficiency of PPM is shown in some practical engineering cases. Sensitivity analysis of input parameters was also performed. The simulation results show that reinforcement corrosion ratio is related to the level of corrosion development and the quality of simulation results depends on the accuracy of statistical parameters from onsite experiments. PPM offers a new way to exhibit the evolution of reinforcement corrosion ratio during its service time.

Keywords: corrosion prediction; path probability model; PPM; dynamic evaluation; durability; steel reinforcement; chloride ingress; steel corrosion; Monte Carlo simulation; probabilistic modelling.

DOI: 10.1504/IJMIC.2008.021164

International Journal of Modelling, Identification and Control, 2008 Vol.4 No.3, pp.268 - 277

Published online: 07 Nov 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article