Evaluation of cyclic bond-slip behavior of smooth bars in low strength concrete: An experimental and stochastic study

Abstract

The vulnerability of non-engineered/substandard reinforced concrete (RC) members could arise from the bondslip failure due to low interaction between the smooth bar and poor-quality (low-strength) concrete. Therefore, the cyclic bond-slip response of plain round bars in poor-quality concrete is characterized by experimental and stochastic approaches. To this end, beam-type bond-slip specimens are first exposed to quasi-static cyclic loading. Then, mathematical models are derived from the experimental hysteric responses. Inherent uncertainties in nominally identical tests reveal the high scatter in the response parameters. Therefore, finite element (FE) models generated by implementing the proposed mathematical expressions for bond-slip response are combined with a suitable stochastic sampling technique, which is Latin Hypercube Sampling (LHS). The apparent outcome is the set of load-displacement curves, which formed an envelope around the deterministic model. The peak capacity is captured satisfactorily by the stochastic-numerical models whereas the overall cyclic response is reproduced less accurately. The evaluated bond stress values and corresponding slips in the confined concrete are higher than the ones in the unconfined concrete. Moreover, the cyclic excitation impacts on the bond strength are more recognizable than monotonic loading. The sensitivity analysis, which grades the contribution of the material parameters on the global response, points out the tensile strength of concrete fct as the most significant parameter while the remaining ones have a weak or no influence.