Probabilistic Identification of Chloride Ingress in Reinforced Concrete Structures: Polynomial Chaos Kalman Filter Approach with Experimental Verification

Abstract

This study presents a Structural Health Monitoring (SHM) framework for assessing the integrity of RC structures subjected to corrosive environmental conditions. The presented framework uses the Polynomial Chaos Kalman Filter (PCKF) for accurate prediction of the stochastic characteristics of the chloride profile in RC structures using real time measurements. The PCKF uses available measurement data of the chloride content at specific locations to update the probabilistic characteristics of the chloride ingress model parameters. These parameters are consequently used to forecast the chloride content profile in RC structures. The work builds on the available literature to quantify the various sources of uncertainty associated with the chloride ingress phenomena. Three long-term experimental data sets are used to assess the efficiency of the presented SHM framework by comparing the framework predictions to real time measurements. The experimental data are also used for sensitivity analysis to highlight the effects of the location and frequency of chloride concentration measurements, as well as the chloride ingress modeling assumptions, on the long-term performance of the SHM framework. The results emphasize the robustness of the presented PCKF approach. PCKF is found able to predict, with reasonable accuracy, the experimental measurements of the chloride content in all data sets.