Experimental and analytical assessment of fatigue damage in reinforced concrete tension members

Abstract

In this paper, the effects of bond deterioration due to fatigue loading on the structural response of reinforced concrete tension members are studied. A comprehensive experimental campaign on reinforced concrete ties and pull-out specimens was carried out under monotonic and repeated loading. The use of distributed fibre optical sensing to assess the local strains allowed the bond deterioration mechanism under repeated loading to be observed. Previously applied load cycles caused a reduction in tension stiffening for load levels below the maximum cyclic load but did not significantly influence the deformation behaviour of the member near the ultimate load. The bond stresses at the maximum cyclic load decreased fairly linearly with the logarithm of load cycles, with the most pronounced reduction in bond stresses near the cracks. At the minimum cyclic load, the bond stress distribution along the tie changed and bond stresses reversal occurred close to the cracks. This caused negative tension stiffening in some cases, i.e., the mean strains exceeded those occurring in a bare reinforcing bar at the same load. The mean steel stresses and the steel stress amplitudes between the cracks increased with the number of load cycles due to the gradual reduction of the concrete contribution and bond degradation. Bond degradation occurred linearly on a semi-logarithmic scale, as expected for the generated bond stresses, which were below the fatigue bond strength. The test results highlight that accounting for shrinkage-induced strains is essential for a realistic evaluation of tension stiffening and residual deformations in particular.Based on the experimental observations and previously developed models for monotonic and cyclic loading, two approaches were used to predict the response of reinforced concrete tension members under repeated loading: the Local Fatigue Damage Model (LFDM), based on the local bond stress-slip relationship, and the adapted Tension Chord Model (TCM) which assumes constant bond stresses and mean strains. The results of both approaches agree well with the test results at the maximum load but yielded less satisfactory predictions at the minimum loading level where only the LFDM was able to be applied successfully. Further research is required to validate the negative bond stresses occurring upon slip reversal in the LFDM and the reduction factor proposed in the adapted TCM.