Reinforced concrete confinement coefficient estimation using soft computing models

Abstract

Infrastructure vulnerability toward seismic lateral loading within high seismicity has received massive attention by structural engineers and designers. This is for the purpose to provide a reliable alternative material that strengthening the bending and shear of slabs, columns and reinforced concrete (RC). Despite the utilized approaches of strengthening the concrete structure based on fiber reinforced polymers (FRP) is considerably new technique, exploring more reliable and robust methodologies is the motive of scholars for better structural behaviour understanding. In the current research, two soft computing models called artificial neural network (ANN) and support vector regression (SVR) are applied to predict lateral confinement coefficient (Ks). The models are developed based on gathered dataset from open source researches for the lateral confinement coefficient of columns wrapped with carbon FRP (CFRP) and their corresponding parameters including column width, length and thickness (b, h and t mm), column radius (r mm), compressive strength of concrete (f_c^') and elastic modulus (EFRP). Results indicated the superiority of the ANN model for predicting Ks over the SVR model. The application of the soft computing showed an optimistic approach for the structural lateral confinement coefficient determination.