New empirical formulations for indirect estimation of peak-confined compressive strength and strain of circular RC columns using LGP method

Reinforced concrete (RC) columns have been basically designed to withstand compressive loads by means of strain and ductility of the longitudinal and transverse reinforcing materials. The objective of this paper is to propose new predictive models of confined compressive strength and strain at confined peak stress of circular-reinforced concrete columns using a powerful evolutionary-based computational technique, namely, linear genetic programming (LGP). For this aim, a collection of data is utilized to develop new models. The models obtained in this study characterize peak-confined compressive strength and corresponding strain factors in terms of the compressive strength of unconfined concrete cylinder specimens, core diameter of circular column, yield strength of transverse reinforcement, ratio of volume of lateral reinforcement to volume of confined concrete core, spacing of lateral reinforcement or spiral pitch, and ratio of longitudinal steel to area of core of section in addition to the column height. These factors have also been considered as the most significant input variables in several models proposed by scholars in the existing literature for approximation of the peak-confined compressive strength and corresponding strain of RC columns. To evaluate the validity of the obtained models, several analyses are conducted and the results are compared with those provided by other researchers to validate and verify the capability of the proposed models. Consequently, the results explicitly approve that the proposed models are of a notably better performance than the traditional models in the literature.