The stress ratio has a significant effect on the fatigue crack growth rate (FCGR). The traditional stress ratio normalization model has problems due to incomplete consideration of correction coefficients and an unsatisfactory normalization effect at high stress ratio.
This paper proposes an improved stress ratio normalization model. By correcting the slope and translation intercept of the FCGR curves at different stress ratios, the unknown parameters were solved and optimized by combining them with the genetic algorithm (GA). Under the small range yield condition, the scope of application of the improved model was extended to the range of elastic-plastic mechanics.
Titanium alloys, aluminum alloys and steels were used as research objects. The effects of stress ratio normalization of traditional and improved models were compared in the range of linear elastic mechanics. The improved model can effectively gather the FCGR data at high and low stress ratios on the same straight line. The coefficients of determination R2 of the linear fits to the data were all raised to 0.98 after normalization. The improved models in the elastic-plastic mechanics range exhibited the same normalization effect. The coefficients of determination R2 were 0.98, 0.97 and 0.98, respectively.
The results show that the improved model based on the GA exhibits good stress ratio normalization effect and wide applicability. It provides a new theoretical basis and reference for variable amplitude fatigue crack growth (FCG) life prediction.
