Artificial neural network and genetic programming for predicting the bond strength of GFRP bars in concrete

Abstract

The bond strength between GFRP bars and concrete is one of the most important aspects in reinforced concrete structures and is generally affected by several factors. In this study, experimental data of 159 notched, hinged, splice and inverted hinged beam specimens from an existing database in the literature were used to develop artificial neural network (ANN) and genetic programming (GP). The data used in modeling are arranged in a format of seven input parameters that cover the bar position, bar surface, bar diameter (d b), concrete compressive strength (f c), minimum cover to bar diameter ratio (C/d b), bar development length to bar diameter ratio (l/d b) and the ratio of the area of transverse reinforcement to the product of transverse reinforcement spacing, the number of developed bar and bar diameter (A tr/snd b). The MAE of testing data was found to be less than 1.06 and 0.76 MPa for the proposed ANN and GP models, respectively. Moreover, the study concluded that the proposed ANN and GP models predict the bond strength of GFRP bars in concrete better than the multi-linear regression model and existing building code equations. A parametric analysis was also conducted using the developed ANN and GP models to establish the trend of the main influencing variables on the bond capacity. Many of the assumptions made by the bond design methods are predicted by the developed models; however, few are inconsistent with the developed models’ predictions.