Theoretical Modeling and Experimental Verification of the Acceleration Factor for Narrow-Band Random Vibration Fatigue Tests.

In conventional vibration fatigue acceleration testing, the two-parameter S-N curve is widely used. However, its inability to accurately characterize the fatigue properties in high-cycle regime introduces intrinsic bias in long-term life prediction, particularly for vibration fatigue dominated by high-cycle failure. To improve physical accuracy, this study focuses on narrow-band random vibration and introduces a three-parameter S-N curve to better capture high-cycle fatigue properties. Based on this, a high-precision analytical model for the vibration fatigue acceleration factor is developed, and a closed-form solution in a generalized inverse power-law form is derived. Experimental validation is performed through vibration fatigue tests on selective laser melting (SLM) aluminum and titanium alloys. The results show that when notable differences exist between two- and three-parameter S-N curves, the three-parameter model significantly improves prediction accuracy of the vibration fatigue acceleration factor, confirming the necessity and effectiveness of this study.