Uncertainty Quantification of Semi Destructive Testing for chloride content assessment for a concrete bridge in maritime environment

Abstract

Harbors play a major socio-economic role in almost all littoral countries with more than 80% of trade exchange and play also usually a key role for military defense. Reinforced Concrete (RC) structures in harbors are designed for a long time (50 to 70 years). Marine environment is considered to be aggressive for RC structures. The main degradation mechanism is chloride-induced corrosion. To prevent any critical failure and optimize the maintenance, a diagnosis is usually made several times during the life-time. It usually relies on inspections which consist usually in Semi-Destructive Testing (SDT) with large uncertainties. Uncertainties may lead to a bad diagnosis, and bad decisions as a consequence. Thus, their quantification is required[bonnet1] . This paper aims at providing the distribution of errors of measurement from on-site assessment of chloride profiles and data treatment from 3 operators.

It is shown that first the error of assessement depends on the chloride level for low contents of chloride. Then we show that this error affects highly (30 to 50% of the coefficient of variation) both Diffusion coefficient and Surface chloride content, the two parameters of Fick Equation.