Reliability Test Tailoring for Reducers Validation in E-Drive Systems

In the rapidly evolving landscape of electrified powertrains, Valeo Power division has emerged as a key supplier of eDrive systems for pure electric and hybrid vehicles. A critical challenge in this domain is translating customer-defined reliability targets and extensive operational duty cycles into efficient and representative validation tests. These duty cycles, encompassing diverse driving conditions and driver severities over a vehicle's lifetime, necessitate a robust methodology for identifying and addressing potential failure mechanisms within the mechanical transmission systems (reducer or gearbox).
This paper outlines Valeo's approach to tailoring reliability tests for reducers, aiming to achieve customer specifications with a limited number of test devices and optimized test durations. The methodology involves several key steps: condensation of real-world drive cycle into a synthetic drive cycle with a reduced duration, test acceleration, stress-strength interference analysis, and statistical assessment using cumulative binomial law.
A crucial aspect of this process is the Physics-of-Failure approach: identification of relevant failure modes and the application of appropriate condensation techniques. For reducer fatigue failure mechanism, particular attention is given to two primary loading types: absolute torque/speed loading (surface damage) and torque inversion cycles (alternate damage). The application of materials fatigue modelling (the Wohler inverse power law and its coefficients), which quantify the relationship between load level and component lifetime, is fundamental to both the condensation process and the overall reliability analysis. This comprehensive approach ensures that the derived test definitions are not only effective for physical validation but also provide valuable insights for advanced simulation and design optimization.

Keywords: reliability, reducer, duty cycle, failure mechanism, condensation, stress-strength, binomial law, test acceleration, simulation.