Root Cause Analysis and Performance Optimization of Drive Axles: A Systematic Approach to Noise Reduction

The technological evolution toward low-noise powertrains in the industrial and construction machinery sectors has significantly increased the acoustic requirements for mechanical transmission systems. As modern engines become quieter, drivetrain-generated noise that was previously masked has become more perceptible, often leading to customer dissatisfaction. This thesis investigates a recurring acoustic anomaly, commonly referred to as gear whine, observed in the F068 drive axle series supplied by Comer Industries to a Dutch company. To systematically address the issue, the research adopts the Six Sigma DMAIC (Define-Measure-Analyze-Improve-Control) methodology, enabling a transition from reactive troubleshooting to a structured, data-driven process optimization approach. During the Define phase, a Pareto analysis of field claims identified noise as the dominant failure mode, particularly affecting units assembled on the Indumec non-automated assembly line. Subsequent forensic teardown activities and high-precision metrological inspections were conducted to characterize the geometric and functional conditions of the returned gear sets. The analytical phase revealed that the acoustic anomaly was closely associated with increased Static Transmission Error (STE), which represents the primary kinematic excitation responsible for gear whine. Further investigation traced the origin of this phenomenon to a combination of supplier component variability and non-optimized mounting conditions within the axle assembly. To mitigate these effects, an improvement strategy was developed involving three main actions: the activation of a structured supplier corrective process based on the 8D methodology, the introduction of a controlled run-in procedure to stabilize the gear contact conditions through Gleason Single Flank Testing, and the optimization of mounting distances to ensure proper backlash and meshing behavior. The effectiveness of these measures was validated through a pilot batch of ten units. The results demonstrated that maintaining the backlash within a range of 0.32–0.54 mm and enforcing a maximum Static Transmission Error threshold of 75 microradians ensures consistent acoustic performance and eliminates the previously observed noise anomalies. In addition, the improvement was also supported by the actions implemented at the bevel gear supplier, where additional quality checks were introduced to ensure that the components complied with the required specifications. Finally, the Control phase consolidates these improvements by updating internal technical specifications and proposing the integration of in-line vibration analysis (FFT) at the End-of-Line testing station, enabling early detection of potential transmission irregularities. Overall, this work demonstrates how the integration of advanced kinematic metrics with a structured continuous improvement framework can significantly enhance the acoustic reliability and competitiveness of modern mechanical drivetrains operating in increasingly silent environments.