Discussion for Predictions of Gigacycle Fatigue Strength in High-strength Steel

Predictions of gigacycle fatigue strength in high-strength steel were discussed by using reported fatigue test results. The discussed prediction models were Tanaka-Akiniwa model, Murakami's equation and a new model. The prediction based on the Tanaka-Akiniwa model overestimated the effects of inclusion sizes, resulting in much lower fatigue strength for a large inclusion than actual. The prediction based on the Murakami's equation provided us good estimations for the effects of the inclusion sizes. However, estimated ODA sizes, which were necessary in the calculation, were imaginary since the estimated fatigue strength did not agree with the fatigue test results. The new model, proposed in this research, was a modification of the Tanaka-Akiniwa model. The new model used a new fatigue crack growth law instead of conventional Paris law. In this new model, coefficient of determination, R-2 value, in the fitting of the fatigue test results was maximized to determine a new constant. The prediction based on the new model provided us much better estimations than that based on the Tanaka-Akiniwa model, and the new fatigue crack growth law showed good agreements with the measured crack growth rates for internal small cracks. Validity of the new model was thus confirmed in this research.