ACI 447.1R-18 Report on the Modeling Techniques Used in Finite Element Simulations of Concrete Structures Strengthened Using Fiber-Reinforced Polymer (FRP) Materials.
3.5—Effect of bond between steel reinforcement and concrete The bond between the steel reinforcement and concrete needs to be carefully considered in any FE model, as it directly infuences the force transferred from the steel bar to the concrete or vice versa. Although perfect compatibility between the steel and the concrete has been assumed by some researchers, the connection between the steel and the concrete is more accurately represented in terms of a peak bond stress and corresponding slip. In the case of longitudinal reinforcement, researchers have found that the assumption of perfect bond between the steel reinforcement and the concrete results in a more distrib- uted diagonal crack pattern than when a bond relationship is assumed (Chen et al. 2012). The efects of distributed cracking are seen as lesser crack widths and a delay in the debonding failure load in the case of FRP-strengthened beams. The assumption of perfect bond and the subsequent reduction of crack widths can also result in higher shear retention and an artifcial increase in the shear capacity of the beam. Two diferent methods have been used to include the bond-slip efect in the FE analysis of RC structures. The frst method is to use a contact element with a predefned bond-slip law between the discrete reinforcement elements and the concrete. The second method is to apply to a bond- slip behavior using one-dimensional bond link elements or spring elements at the nodes connecting the reinforcement to the concrete (Youai 2000).
In contrast to longitudinal steel reinforcement, researchers have found that the predicted load displacement response for RC beams was less sensitive to the assumed bond condi- tion between the steel stirrups and concrete. Although many existing studies conducted in the area of fexural strength- ening of RC beams adopted an accurate bond-slip model between the internal steel stirrup reinforcement and the concrete; the practice was less common for shear strength- ened beams. Only a single study by Chen et al. (2012) was found to investigate alternative bond-slip models between internal steel stirrup reinforcement and surrounding concrete. Both weakly bonded, strongly bonded, and perfectly bonded scenarios were investigated to represent the bond behavior between the internal steel stirrups and the concrete. When examining the efects of the assumed bond condition with respect to the peak load predictions, the weakly bonded stirrups provided an underestimation of the failure load while a strongly bonded scenario provided good predic- tions that matched the experimental data in terms of peak load and ductility. The assumption of perfect bond resulted in a similarly good prediction of peak load, but produced greater post-peak ductility behavior, which was absent from the experimental results. Furthermore, it was found that the bond condition afected the crack pattern, and the process of crack propagation in a complex manner and further investi- gation was required to achieve a better understanding.ACI 447.1R pdf download.