Using genetic programming to model the bond strength of GFRP bars in concrete under the effects of design guidelines

Abstract

Purpose This paper aims to use a derivative of genetic programming to predict the bond strength of glass fiber-reinforced polymer (GFRP) bars in concrete under the effects of design guidelines. In developing bond strength prediction models, this paper prioritized simplicity and meaningfulness over extreme accuracy. Design/methodology/approach Assessing the bond strength of GFRP bars in concrete is a critical issue in designing and building reinforced concrete structures. Findings Ultimately, the equation of a linear form of a particular design guideline was suggested as the optimal prediction model. Improvements to the current design guidelines suggested by this model include setting a 1.31 magnification and considering the effects of the three significant parameters of bar diameter (db), minimum cover-to-bar diameter (C/db) and development length to bar diameter (l/db) under an acceptable root mean square error accuracy of around 2 MPa. Furthermore, the model suggests that the original influence parameter of concrete compressive strength (fc) may be removed from bond strength calculations. Originality/value The model suggests that the original influence parameter of concrete compressive strength (fc) may be removed from bond strength calculations.