Influence of steel-concrete interface defects owing to the top-bar effect on the chloride-induced corrosion of reinforcement

This paper deals with the influence of the steel-concrete interface condition on the reinforcement corrosion in a chloride environment. Two large-sized concrete members were cast including horizontal reinforcements at different levels in order to create voids under the upper horizontal bars caused by the well-known 'top-bar effect'. The walls were then sawn into small-sized specimens including one rebar. These small specimens were submitted to wetting-drying cycles in salt water (natural corrosion process). The polarisation resistance technique was performed to monitor the steel corrosion during the ageing up to 55 weeks. Chloride profiles were regularly measured and specimens were broken to observe the corrosion pattern. Results show that, first, a good-quality steel-concrete interface significantly reduces the corrosion rate in comparison with the specimen with interfacial defects. Second, before concrete cracking, the corrosion pattern observed is completely different between the specimens with or without defect. Perfect interface specimens show a typical local pitting corrosion pattern owing to chloride attack. However, specimens affected by the 'top-bar effect' show a generalised corrosion along the steel surface affected by the void (surface under the horizontal steel bars in regard to concrete casting direction). Macrocell corrosion seems to develop between the lower surface affected by the void acting as an anode and the upper surface where the bond is perfect as the cathode. The concrete cracking is delayed for specimens affected by the 'top-bar effect' in spite of a higher corrosion rate. This delay can be attributed to the time necessary for the rust to fill the voids.