Abstract
This research studies the structural performance of glass fiber-reinforced polymer (glass-FRP) reinforced concrete (RC) columns incorporating glass hybrid fibers (GFC) and steel rebars RC columns having steel hybrid fibers (SFC) using experiments, artificial neural networks (NN), and theoretical modeling. A set of 18 circular concrete columns, each with a diameter of 300 mm and a height of 1200 mm, was constructed and axially loaded to failure. Glass fibers and steel fibers were incorporated to produce hybrid fiber-RC (HFRC). Nine samples were manufactured with glass-FRP reinforcement, while the remaining nine were manufactured with steel reinforcement. According to the findings, the GFC columns had lower axial strengths (ASs) up to 20%, and higher ductility indices up to 26% than the SFC columns. Both GFC and SFC columns showed the same influence of eccentric loading in the form of a decrease in AS. To develop a novel NN model, a database of 275 specimens of glass-FRP-RC columns was gathered from previous studies. To achieve an optimum model, the NN model was calibrated for different numbers of neurons in the hidden layers (HLs). A novel theoretical equation for determining the AS of GFC columns was also suggested. The proposed theoretical and NN models showed average discrepancies of 3.2 and 1.9% from the test results of both GFC and SFC columns.