Dynamic Simulation Method for Probability Evolution of Reinforcement Steel Corrosion

SONG Zhi-gang,JIN Wei-liang,LIU Fang,NI Guo-rong
DOI: https://doi.org/10.3785/j.issn.1008-973x.2006.10.019
2006-01-01
Abstract:The probability density function(PDF) of steel corrosion ratio(SCR) evolving with service time(ST) was simulated to study the performance degradation of chloride-ingress RC member.A path probability method(PPM) was suggested to analyze the PDF of SCR under given ST.In PPM,the whole corrosion process of the given ST is divided into a series of paths constructed by corrosion initiation phase and corrosion expansion phase,the probability of which can be described by the probability of corrosion initiation and the conditional probability of SCR,in condition that the corrosion initiation has happened,respectively.The unconditional SCR's PDF of given ST can be achieved by summing unconditional PDF of all paths after the unconditional PDF of each path is evaluated by Bayesian formula.Based on PPM,the dynamic evolution of SCR's unconditional PDF with increasing series of STs was simulated through Monte Carlo simulation technique.A practical engineering RC member was analyzed to illustrate the evolution results.The validity of the simulation method and the sensitivity of input parameters were studied.The results showed that the distribution parameters of SCR can be effectively evaluated by PPM,and that both the probability distribution types and the distribution parameters change with ST.So that it could induce bias in design if only one probability distribution type is used.
What problem does this paper attempt to address?