Effects of Bond Deterioration Due to Corrosion on Seismic Performance of Reinforced Concrete Structures

Abstract

Reinforced concrete structures deteriorate throughout their lifetime. This is particularly apparent in structures subjected to aggressive environments, which results in corrosion of reinforcing steel. Designers make allowances for accelerated deterioration in these environments in an attempt to ensure the durability of the structure. To combat corrosion, improved concrete characteristics and additional concrete cover are used to increase the protection provided by concrete to reinforcing. In spite of these measures, cracking of structures in service and from natural hazards can limit the effectiveness that these measures provide. Ultimately, this results in structures suffering from corrosion, which affects their strength, stiffness, and ductility. While strength reduction can be associated directly with a reduction in bar area, impacts on stiffness and ductility are associated with more complex mechanisms, one of which is bond deterioration. A key assumption in reinforced concrete design is that there is perfect bonding between steel reinforcing and surrounding concrete to allow for strain compatibility to be assumed. Perfect bond does not exist and diminished bond performance due to corrosion deterioration further violates this assumption, the effects of which are not fully understood. This thesis investigates the effects of bond deterioration due to corrosion on the seismic performance of reinforced concrete structures. 60 monotonic and cyclic pull-out tests were undertaken on corroded reinforced concrete specimens, with corrosion levels ranging from 0% to 25% reinforcing mass loss. Additional tests were also conducted on specimens with variations in the amount of confining steel to simulate losses in confinement associated with corrosion of confining steel. Experimental results were used to develop corrosion and confinement dependent cyclic bond-slip model. The proposed bond-slip model was then used to modelling pull-out of reinforcing bars detailed in accordance with New Zealand design standard NZS3101. Analyses were performed at a range of corrosion levels, levels of confinement, and uncorroded bond strengths. These showed that pull-out of reinforcement occurred at as little as 8% corrosion in low strength, unconfined conditions. Multi-spring modelling of standard reinforced concrete columns, representing a bridge pier to foundation connection, was performed at the full range of deterioration with allowance for bond slippage. These analyses showed significant reductions in stiffness occurring with increased corrosion levels as well as reduced ductility and possible pull-out of reinforcement.