Anchorage of naturally-corroded, plain bars in Reinforced Concrete structures

Licentiatavhandling, 2019

Reinforced Concrete (RC) is the most common construction material in existing structures. However, RC structures are susceptible to deterioration over time, with corrosion of the reinforcing steel as most common mechanism. Corrosion reduces strength and ductility provided by the reinforcement bars and affects their interaction with the concrete. Research on the structural effects of corrosion commonly focuses on deformed bars and applies artificial corrosion. Performance is evaluated based on testing the bond between the bars and the concrete. Plain (smooth) reinforcing bars, as typical in older structures, are seldom studied. Plain bars interact differently with concrete, due to the absence of ribs. Additionally, doubts on the relevance of artificial corrosion methods have been raised. Thus, there is a lack of knowledge on the effects of corrosion of plain bars, and naturally corroded specimens are the ideal mean of acquiring it.
This work investigates the bond of naturally corroded, plain reinforcement bars by testing specimens taken from the edge beams of a decommissioned, 80-year-old bridge. Pilot tests were performed to investigate possible test configurations, to which a total of 20 beams were subjected to displacement-controlled 3-point bending. The beams presented different levels of damage, and the corrosion level of each tensile reinforcement bar was afterwards measured using of a 3D scanner. All but three of the tested specimens were able to anchor the yield force of the bars after the opening of one or two major bending cracks in the middle. Bending failure, not bond strength, limited the load-carrying capacity for the majority of test specimens. At large deflections, end-slip of the reinforcement bars was observed; thus, anchorage limited the deformation capacity.
The average bond strength was evaluated separately in the unyielded and in the yielded zones. The average bond strength in the unyielded zone was found to be equal to 7.4 MPa, with a standard deviation of 3.3 MPa. The casting position was identified as an important factor. Bottom-cast bars had higher bond strength when uncorroded, but were more prone to external cracks in the bond region and consequently loss of bond strength for small corrosion levels. Top-cast bars had lower bond strength when uncorroded, but reached higher bond strength with increasing corrosion levels, due to the absence of external cracks. These differences are likely the result of a higher density of the concrete surrounding the bottom-cast bars. In the yielded zones, substantial loss of bond strength was observed, with an average of 1 MPa. This affected the overall structural behaviour, which was observed to change from beam to arch action for larger deflections.
To conclude, the results improve our understanding of the behaviour of older structures with plain bars and will enable, in the long run, improved assessment methods.