Time-dependent behaviour of reinforced concrete beams under sustained and repeated loads

This thesis primarily aims to explore the flexural behaviour of reinforced concrete beams subjected to different loading types. Both bonded and partially debonded (in the maximum moment zone) reinforced concrete beams under static sustained and repeated loading were investigated. Information relating to surface strains and mid-span deflection were continuously recorded for a period of 90 days so that meaningful comparisons could be made between the structural displacements of the beams tested under different load levels. The range of the sustained load applied varied from that corresponding to the first cracking moment to that required to produce a stabilised crack pattern. The experimental outcomes show that the long-term mid-span deflection of the reinforced concrete beams where the reinforcement was artificially debonded from the concrete is substantially higher than that of bonded reinforced concrete beams under sustained loading. For beams subjected to repeated loading, the amplitude of the repeated loading was deemed to be around one eighth of the sustained load. The bond between concrete and steel in reinforced concrete beams subjected to a repeating load can be significantly damaged due to the loading even though the frequency is relatively low (i.e. 0.2 Hz). On the other hand, and more unexpectedly, for the cyclically applied loading there was no substantial difference between the observed ultimate deformations of the bonded and debonded beams. Moreover, there is a linear relationship between the number of cracks and the shrinkage deflection. Beams having a higher number of cracks develop more deflection due to shrinkage. Nonlinear finite element software (Midas FEA) was used to simulate the experimental tests. It was found that a numerical-experimental match could only be achieved after applying necessary modifications to the shrinkage strain distribution along the beam section. In addition, the capacity of the software to separate the shrinkage and creep deflection clearly allows the relationship between number of cracks and shrinkage to be observed, and confirms what was observed in the experimental investigation.