Two-stage deterioration model updating of RC structures in marine environment using long-term field inspection data

Abstract

For the deterioration of reinforcement concrete (RC) structures in marine environment at various stages, models on chloride penetration, crack initiation and crack propagation have been proposed and studied in literature. However, to be applied to real existing structures, these models, usually associated with significant uncertainties, have to be calibrated and modified with long-term in-situ inspection data. Toward this goal, this paper develops a two-stage updating framework to incorporate measured data into existing models to achieve a more realistic modeling of deterioration process. To improve the computation efficiency, the framework can be integrated with simple rejection sampling and other advanced sampling techniques such as the modified sequential adaptive importance sampling (SAIS). A real case study of high-pile wharf structure in Guangzhou Port is carried out to demonstrate the updating effect of the proposed framework. The observed data of chloride content profile and steel bar corrosion depths from one field inspection are used to construct the likelihood functions in the two updating stages. Comparison between the results of the first and second updating stages is made, and the sensitivities of different parameters to the data are analyzed, subsequently the year of reinforcement corrosion initiation and the evolution of corrosion depth are inferred using these results. At last, the data from an early inspection are also incorporated to investigate the impact of two inspections on updating effects.