Theoretical Model of Seismic Fracture Prediction for Low-Yield-Point Steel LYP225 and Its Validation

The fracture of energy-dissipation components with low-yield-point (LYP) steels under seismic action may cause the progressive collapse of the structures. Therefore, it is essential to predict the fracture of various LYP steel energy-dissipating components under the seismic action and ultra-low cycle fatigue loading. To address this issue, a seismic fracture prediction method for LYP225 steel subjected to cyclic loading was proposed and validated. First, based on updated Voce–Chaboche model and small strain assumption, an improved Voce–Chaboche (IVC) model was proposed to reasonably characterize the stress–strain relation of LYP225 steel under cyclic loading. Five cyclic coupon tests were conducted to validate the proposed IVC model. The proposed IVC model is foundational for accurately predicting the fracture of LYP225 steel under cyclic loading. Second, a general fracture model accounting for the effects of stress states, linearity, nonlinearity, and loading history was developed to predict the fracture of LYP225 steel under cyclic loading. Twenty cyclic tests with various notch details were conducted to validate the proposed fracture model. The experimental results were presented and analyzed in detail, including from the results of scanning electron microscopy. Third, the calibrated fracture model can accurately predict the fracture index and cumulative displacement of fracture initiation point, with an average accuracy of 99.3% and 95.3%, respectively. The proposed fracture model was incorporated into the developed material subroutine to simulate the steel failure. Numerical results were in good agreement with experimental results, and fracture initiation points for all specimens can be reasonably predicted. Finally, the proposed IVC model, fracture model, and their numerical implementation would contribute to reasonably predict the potential failure of steel members, connections, and structures under the seismic actions, fatigue, and cyclic loadings. Due to the high prediction accuracy and practicality of the proposed fracture model, it is recommended to integrate it into finite-element software and use it for engineering applications. This action also would contribute to reduce or avoid potential casualties and property losses in earthquake disasters.