Crystal plasticity modeling of low-cycle fatigue behavior of an Mg-3Al-1Zn alloy based on a model, including twinning and detwinning mechanisms

The low-cycle fatigue of strongly-textured Mg alloys is dominated mainly by an alternation of twinning-detwinning process, which poses a great challenge to conventional constitutive models. As a result, a good reproduction of both mechanical response and deformation behavior often fails in previous simulations, in particular for high strain amplitudes. In the present study, the numerical simulations of low-cycle fatigue behavior of an Mg AZ31 plate with a +/- 2% axial strain amplitude along the rolling direction were conducted, using a modified crystal-plasticity based on the finite strain elastic-viscoplastic self-consistent model containing twinning and detwinning mechanisms (EVPSC-TDT), and the results were compared with the mechanical response and deformation behavior determined by a real-time in-situ neutron diffraction. The simulations well predict the hysteresis loops and cyclic-hardening responses throughout the whole fatigue life (80 cycles), along with the evolution of the maximum twin volume fraction and residual twins. For the first time, the evolution of lattice strain throughout the whole fatigue life was calculated, and the results are in good accordance with that determined by neutron diffraction. The present study strongly demonstrates that the EVPSC-TDT model is very effective for modeling the low-cycle fatigue behavior of Mg alloys. More details about the evolution of the maximum twin volume fraction with fatigue cycles are found, an increase from 1st to 2nd cycle, followed by a decrease from 2nd to 20th cycle and a secondary increase after 20th cycle. This finding corrects the previous opinion that the maximum twin volume fraction increases at the initial stage and tends to be saturated at the late stage. Combining the results of simulations and neutron diffractions, the corresponding reasons are explained. Afterward, the mechanisms of the cyclic-hardening behavior at tension stage and compression stage are revisited, and some new insights are provided.

Automated summary

This study investigated the low-cycle fatigue behavior of strongly-textured Mg alloys through numerical simulations and in-situ neutron diffraction experiments. The results demonstrate the effectiveness of the used model in simulating the low-cycle fatigue behavior of Mg alloys. Additionally, the study revealed a specific evolution trend of the maximum twin volume fraction with fatigue cycles.