The purpose of this study is to develop two empirical models that predict the shear strength of exterior beam-column joints exposed to monotonic and cyclic loading using Gene expression programming (GEP).
The GEP model developed for the monotonic loading case is trained and validated using 81 data test points and that for cyclic loading case is trained and validated using 159 data test points that collected from different 9 and 39 experimental programs, respectively. The parameters that are selected to develop the cyclic GEP model are concrete compressive strength, joint aspect ratio, column axial load and joint transverse reinforcement. The monotonic GEP model is developed using concrete compressive strength, column depth, joint width and column axial load.
GEP models are proposed in this paper to predict the joint shear strength of beam-column joints under cyclic and monotonic loading. The predicted results obtained using the GEP models are compared to those calculated using the ACI-352 code formulations. A sensitivity analysis is also performed to further validate the GEP models.
The proposed GEP models provide an accurate prediction for joint shear strength of beam-column joints under cyclic and monotonic loading that is more fitting to the experimental database than the ACI-352 predictions where the GEP models have higher R2 value than the code formulations.
