Probabilistic service life modeling of RC structures subjected to the combined effect of chloride-induced corrosion and cyclic loading
Résumé
Durability of reinforced concrete (RC) structures placed in non-aggressive environments is often sat- isfactory. However, under certain environmental conditions there are internal or external actions that significantly reduce their lifetime. Nowadays, lifetime assessment of deteriorating structures is focused on the isolated effect of the main deterioration processes (corrosion, fatigue, creep, etc.). However, it is paramount to study the combined effect of various deterioration processes because such interaction could reduce structural integrity. This study proposes a new model for lifetime assessment of RC structures subjected to corrosion- fatigue deterioration processes. The proposed model distinguishes between the following phenomena: chloride penetration, corrosion of reinforcement, concrete cracking and corrosion-fatigue of reinforcing bars. Chloride penetration determines corrosion initiation. Corrosion reduces the cross-section of reinforcing steel. The accumulation of corrosion products in the steel/concrete interface produces concrete cracking. Fatigue causes the nucleation and the propagation of cracks in steel bars. The interaction between corrosion and fatigue can only be taken into account when modeling the combined problem. Thus, pitting corrosion generates stress concentrations that nucleate cracks in the reinforcing bars. Cyclic loading and environmental factors affect the kinematics of crack propagation. There exists significant uncertainty related to the combined corrosion-fatigue problem. This study also addresses this problem by considering the uncertainty inherent in (1) material properties, (2) model and its parameter and (3) environmental actions. Time-invariant random variables represent the uncertainty of material properties and model. Stochastic processes consider the uncertainty of environmental actions. The proposed stochastic models for weather (temperature and humidity) take into account seasonal variations and global warming. The stochastic models for environmental chloride concentration differentiate between exposure to de-icing salts or the sea. Finally, a fuzzy approach accounts for the uncertainty related to corrosion rate. The proposed model is applied to the reliability analysis of RC members located in various chloride- contaminated environments. Overall results reveal that the combined effect of corrosion-fatigue de- pends on environmental conditions and strongly influences the performance of RC structures leading to large reductions in expected lifetime.
Domaines
Génie civil
Loading...