Improved shear strength model for exterior reinforced concrete beam-column joints using gene expression programming

Abstract

Beam-column joints without transverse reinforcement can be considered as a weak link in reinforced concrete (RC) moment-resisting framed structural systems, particularly when subjected to seismic lateral loads. It is essential to fully understand the shear behavior of RC beam-column joints to avoid significant damage and eventual failure of RC structures in the case of joint panels excessively damaged. This study aims at developing a framework for systematic understanding of the shear behavior of RC beam-column joints where a novel joint shear strength model is presented. Totally, 56 test results on exterior RC beam-column joints without transverse reinforcement from various sources were used. The proposed model was generated by the utilization of the gene expression programming (GEP) approach. Thus, statistically, significant tests were applied for evaluating the different parameters influencing the shear strength of the joint, e.g., material features, design variables, and joint geometrical and detailing configurations. The proposed model was able to appropriately reflect the contribution of each variable of the beam-column joints involved in the GEP. The presented findings also include a comparison of the developed GEP-model with the existing analytical models, i.e., the conjectures of these available models were compared with the achieved results of this study. It is demonstrated that the presented GEP-model more accurately predicted the shear strength of RC beam-column joints than the existing models, having a much smaller coefficient of variation of 0.093 for the ratio between the predicted and experimentally determined shear strength values, compared to other models available in the literature.