Proceeding of the First International Seminar on Materials and Engineering Structures (ISMES’2024)

Abstract

In the last two decades, several studies have been performed to investigate the behavior of concrete columns internally strengthened with glass fiber-reinforced polymer (GFRP) bars. Consequently, numerous models have been suggested to predict the axial load-carrying capacity (ALCC) of the confined compression elements. All of the ALCC models available in the literature have been developed based on small number of data and general regression method with limited variables of such column elements. The use of the artificial intelligence (AI) is one of advanced technique for precisely estimating the behavior of composite structural elements by considering a large number of variables. The aim of the present study is to propose a new model for the ALCC of GFRP-reinforced concrete columns using Random Forest Algorithm. To achieve this objective, a large number of experimental data of 235 GFRP-reinforced concrete columns have been collected from previous experimental studies. A statistical assessment of eight ALCC models has been conducted using three statistical indices namely; coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) against of the collected database. The obtained results show that the proposed model using Random Forest Algorithm give the highest performance compared with the eight existing models with R2 = 0.97, RMSE = 567.97 kN, MAE = 208.83 kN. Consequently, the proposed model can accurately predict the ALCC of FRP-reinforced concrete columns that can be used for the design of such compressive elements.