Fatigue Properties and Life Prediction of GS80A Steel Under the Effect of Hydrogen-Rich Environment

To ensure the safety of high-strength bolt materials in an ocean hydrogen-rich environment, this work first studies the H-effect on GS80A steel by mechanical tests and fracture surface observations. The experimental mechanical properties with/without the H-effect are compared with each other, showing a strong hydrogen-induced reduction of ductility, ultimate tensile strength, yield strength, and full-cycle fatigue life. To reproduce the S-N curve based on the experimental fatigue life with/without the H-effect, a neural network model and a typical gray theory are used in this paper for the first time. Reasonable agreement is found between these two models. Finally, the H-affected P-S-N curves are obtained based on the Bootstrap method, indicating the fatigue life obtained by the same condition convergence to Weibull distribution.

Automated summary

In this work, the influence of hydrogen on GS80A steel was studied through mechanical tests and fracture surface observations. The results showed a significant reduction in ductility, ultimate tensile strength, yield strength, and fatigue life due to hydrogen. For the first time, a neural network model and a typical gray theory were used to reproduce the S-N curve based on experimental fatigue life with and without the presence of hydrogen, and reasonable agreement was found between the two models. The H-affected P-S-N curves were obtained using the Bootstrap method, indicating that the fatigue life obtained under the same conditions converged to a Weibull distribution.