New empirical approaches for compressive strength assessment of CFRP confined rectangular concrete columns

Utilizing of CFRP for confining concrete columns has been demonstrated to improve the capacity and ductility of columns. In reason of numerous parameters that affect the compressive strength of the CFRP confined rectangular concrete columns (CRCC), developing formula is complicated. In the current study, three methods are evaluated for predicting the residual compressive strength of CFRP CRCC. Multiple regressions (MR), stepwise regression (SR), and artificial neural network (ANN) models were extended as credible methods for generating and evaluating the compressive strength of CFRP CRCC. The required data for training algorithms obtained from a reliable database. The accuracy of the developed formulae is verified using appropriate criteria. Then, a comparison was made between proposed formulae-based models to examine the accuracy degree of these methods. It is understood that despite a negligible difference between proposed models results the obtained formulae based on the SR and ANN methods give the exact results than the MR model.