Abstract:The bond strength between rebar and concrete is a parameter that affects the bearing capacity of reinforced concrete members. Corrosion can impact the bond strength between rebars and the surrounding concrete negatively, so its accurate determination, especially when the rebar is corroded, is an important aspect in studying the structural performance. This study first created a comprehensive database by collecting 482 experimental data related to the bond strength of corroded rebars (found from previous studies’ pullout and beam tests) to predict the ultimate and relative bond strength between corroded rebars and surrounding concrete with/without transverse rebars using such machine learning methods as the multi-layer perceptron, radial basis function neural networks, and support vector regression. Next, some empirical models were collected from previous studies that could determine the relative bond strength. The results of these models and the mentioned methods were compared, which indicated that the experimental models had low accuracy, while machine learning methods were highly accurate.

Predicting the Ultimate and Relative Bond Strength of Corroded Bars and Surrounding Concrete by Considering the Effect of Transverse Rebar Using Machine Learning