Quantification Of Condition Indicator Performance On A Split Torque Gearbox Using An Advanced Signal Processing Technique

The advent of Split Torque Gearboxes (STG) has provided a low weight, high reduction ratio alternative to the traditional planetary system used as the last the stage of gear reduction to the main rotor. The STG, while a good design, poses significant challenges to the current suite of gear analysis used in Health and Usage Monitoring Systems (HUMS). Traditionally, gear algorithms have used the signal processing techniques of time synchronous average (TSA) and fast Fourier transforms (FFT) to separate gears in the frequency domain that are close to a given sensor. However, given the design of the STG and the large number of other rotating components in close proximity to the sensor measuring the signal, there is a significant reduction in the signal to noise ratio (SNR). The reduction in SNR in turn reduces the effectiveness of the gear algorithms. In this study, a STG was developed to generate vibration data for a healthy gear and one with 50% tooth loss. The gears were run at 0%, 50% and 100% loading conditions. The vibration data collected from the test stand was processed using an advanced signal processing technique called Narrowband Interference Cancellation (NIC). Condition indicators (CI) were computed and responsiveness to fault was measured. The measure of fault responsiveness was separability.