| Abstract: | The aim of the present study is to propose innovative predictive models for shear capacity of reinforced concrete (RC) exterior joints in terms of multiple soft computing techniques. Existing models were evaluated and by a preliminary sensitivity analysis, seven parameters including compressive strength of concrete, product of the yield stress and the reinforcement ratio of the joint stirrups, the effective width of the joint panel, cross‐sectional column width, beam tensile longitudinal reinforcement ratio, beam compressive longitudinal reinforcement ratio, and column longitudinal reinforcement ratio were considered. Then, a large data set having the details of experimental programs on shear capacity of exterior RC beam–column joints was provided. The experimental data were utilized in developing the proposed models. After verification of the new models against available database, their efficiency compared with existing models was confirmed. Finally, a sensitivity analysis was performed in order to find the relative importance of each input parameter on the shear strength of RC joints. The results indicated that the beam reinforcement is the most important factor in shear capacity estimation of exterior RC beam–column connections. |
