Papers »

A polynomial chaos based kalman filter approach for service life prediction of reinforced concrete structures subject to chloride induced corrosion

Chloride induced corrosion is a main factor for the deterioration of reinforced concrete structures. Once chloride concentration at the level of rebar exceeds a critical threshold, corrosion initiates and propagates progressively, degrading the structural resistant capacity and decreasing the design safety margin. Thus, to mitigate risk associated with corrosion initiation and to assign efficient maintenance schedules, accurate prediction of corrosion initiation time and continuous update of the state of the structure are becoming critical for both safety and serviceability proposes.
In this study, the Polynomial Chaos Kalman Filter (PCKF) is employed to update the state of the structure and accurately predict the corrosion initiation time based on embedded sensors measurements of chloride concentration. This recently developed filter demonstrated accuracy and efficiency when incorporated in a non-Gaussian non-linear probabilistic framework. The output of the filter is efficiently implemented in a Monte Carlo simulation scheme to predict the reliability of the structure with time, and to accordingly update maintenance schedules. This data assimilation technique is advantageous when compared to traditional deterministic and probabilistic models, especially with the well-known high uncertainty associated with corrosion initiating time estimation.

Author(s):

Wael Slika    
American University of Beirut
Lebanon

George Saad    
American University of Beirut
Lebanon