Calculation of reinforced concrete corrosion initiation probabilities using the Rosenblueth method

Abstract

The Rosenblueth method is used to analyse stochastic parameters in a numerical model to calculate the solution for independent variables. By replacing all or a part of the independent variables with random variables, the solution becomes itself random. The Rosenblueth method simplifies this passage, while being limited to dichotomous random variables. This avoids the complex and computer-intensive Monte Carlo simulation method. A specific example is developed in this article using a non-linear transport model to simulate chloride ion penetration into reinforced concrete and to predict the development of corrosion in the civil structure. The taking into consideration of complex physical chemical processes and a reconstituted climate increases considerably the simulation time, making impossible the usual use of the Monte Carlo method. The two numerical examples show that the exposure degree affects in a dominating way the appearance of structural damage.