An Improved Giuffré–Menegotto–Pinto Model: Implementation in OpenSees and Verification With Experimental Results

ABSTRACT The Giuffré–Menegotto–Pinto (GMP) model is widely used by researchers and practitioners for simulating the cyclic response of structural and reinforcing steel materials. Filippou et al. incorporated cyclic isotropic hardening into the model and showed that it fails to return to the original loading path after partial unloading, causing stress overshooting. Recently, Bosco et al. proposed a modification to rectify this error. While their modified curve is useful, the proposed approach does not eliminate the overshooting error and can even lead to an unintended impasse that prevents further state determination during nonlinear dynamic analyses of structures subjected to seismic excitations. In a more recent study, Mortazavi et al. showed that the GMP model can lead to a creeping effect towards the end of a seismic excitation due to premature activation of the Bauschinger effect. To address these limitations, a three‐level curve approach is proposed, where the first level curve uses the original GMP model, and the other two use the modified equation of Bosco et al. The fundamental attributes of the proposed modifications are: (i) detection of overshooting, (ii) its elimination through curve selection, and (iii) control of the Bauschinger effect. The improved model is implemented in OpenSees and thoroughly verified using laboratory test data. Nonlinear dynamic analysis results for single‐ and multi‐degree‐of‐freedom systems demonstrate the model's efficacy. Using a tall building model, it is demonstrated that the proposed modifications require no significant additional computational effort for nonlinear dynamic analyses of structures.