GRNN Predictor for joint strain-stress curves

A GRNN model trained by 90 experiments to predict critical strain and stress points of the strain-stress history envelope of an RC joint

A Generalized Regression Neural Network (GRNN) model trained by 90 experiments to predict critical strain and stress points of a reinforced concrete frame joint. The model uses joint type, number of transverse beams, joint concrete strength, horizontal reinforcement yielding strength, joint reinforcement ratio, joint width, beam reinforcement yielding strength, beam reinforcement ratio, height of the beam, width of the beam, column reinforcement yielding strength, column reinforcement ratio, height of the column, width of the column, eccentricity of the beam, axial force ratio and observed failure type as input while the outputs are the strain and stress points at first stiffness loss, maximum strength and ultimate deformation.