Abstract: Splines play an irreplaceable role in mechanical transmission systems, and their life directly affects the reliability of equipment. Most of the existing spline fatigue life prediction models only consider the influence of parameters such as loads and material uncertainty, but ignore the machining errors generated in manufacturing. Machining errors lead to different clearances of the spline, making the force distributed to each tooth different. The damage to the tooth with a large load is more serious, which reduces the spline life. Therefore, to obtain a more accurate spline fatigue life, based on the requirements of the manual for the spline errors, the distribution range of the spline clearance at different machining accuracy is deduced. Then, the effect of the clearance uncertainty on the number of engaged teeth and uneven distribution of force is considered, an equivalent model of the force carried by each tooth is established. Finally, the fatigue life distribution of the spline is obtained, and compared with the spline fatigue life calculated by the traditional method. The case results show that the average spline fatigue life is 10⁸ load cycles, and the probability of fatigue failure in 2×10⁷ to 7×10⁸ load cycles reaches 50%. The proposed probabilistic fatigue life prediction method can further consider the impact of the clearance uncertainty and provide guidance for the reliability design of splines.

Key words: spline, tooth clearance, uncertainty, numerical model, fatigue life