ABSTRACT

During operation, gears usually face variable loads and load sequences. For instance, the load of transmission gearing mainly depends on the input torque, which in many cases is inconstant. These variable loads have to be considered in the design process of gears to achieve an optimal dimensioning of their load carrying capacity.

Standardized methods, like those presented within ISO 6336 or ANSI/AGMA 2101-D04, factor in variable loads when calculating the service life. These methods all trace back to the linear approach published by Miner and Palmgren in the first half of the 20th century. While this approach considers variable loads, it does not take the load sequence into account. Even though this approach delivers good and useful results, research has shown that neglecting the load sequence is a potential issue. In reality, different load sequences can result in an immensely different service life, even though the same service life is to be expected according to the linear approach. Because of this fact, many so-called non-linear approaches have been published. All these non-linear methods take the load sequence into account but come with the tradeoff of a higher complexity. Today, the use of non-linear methods is uncommon for gear-design and there is no validation for their suitability.

In this paper, an extensive set of experimental data investigating the influence of different load sequences on the service life of gears is presented. The data derives from tooth root breakage tests conducted at the FZG. To evaluate the usefulness of non-linear methods for the service life calculation of gears, the data is first analyzed with the linear approach to set a benchmark. In the next step, a non-linear method is applied. Finally, the results of both approaches are compared to analyze if the use of the more complex non-linear method for the service-life-calculation of gears yields potential benefits. 

Author(s): Daniel Vietze, Josef Pellkofer, Karsten Stahl

ISBN: 978-1-64353-153-3