A cyclic GTN model for ultra-low cycle fatigue analysis of structural steels

Ultra-low cycle fatigue (ULCF) is a critical concern in the seismic design of steel structures due to its adverse effects on the ductility and energy dissipation capacity of steel connections. The Gurson-Tvergaard-Needleman (GTN) model, a well-accepted micromechanical fracture model, cannot be applied directly to ULCF analysis, as it requires proper estimation of the effects of cyclic loadings on the ductile fracture of materials. In this paper, a cyclic GTN (C-GTN) model was proposed for the ULCF prediction of materials, in which the evolution of microvoids during ULCF loadings was addressed for tensile and compressive half load cycles, respectively, and the effect of the cumulative plastic strains on the material's resistance to ductile fracture was considered by the reduction of the critical void volume fraction for void coalescence. Then, a VUMAT subroutine was programmed for the C-GTN model, in which the cyclic hardening behavior of materials was simulated with the Voce-Chaboche model. Finally, the C-GTN model was calibrated for the G20Mn5QT cast steel based on previous tests on notched round bar specimens of the material. The applicability of the C-GTN model and the VUMAT subroutine were validated in the ULCF analysis of double-hole plate specimens of the G20Mn5QT cast steel.

Automated summary

This paper proposes a cyclic GTN (C-GTN) model for predicting the ultra-low cycle fatigue (ULCF) of materials. By considering the evolution of microvoids and the effect of cumulative plastic strains on ductile fracture, this model can better describe the behavior of materials. The applicability of the C-GTN model is validated using a VUMAT subroutine.