Gear Rattle Analysis Based on Wavelet Signal Decomposition

Abstract

The “gear rattle” phenomenon is a research topic of great interest for the NVH (Noise, Vibration and Harshness) automotive sector, concerning driveline noise and vibrations coming from the manual gear boxes. It is due to the internal combustion engine variable torque producing impacts and rebounds, and consequently noise, between the teeth of the unloaded gear pairs of the gear box because of the unavoidable presence of backlashes. Discrete Wavelet Transform (DWT) is used to decompose the angular relative motion signal and the wavelet decomposition details are adopted to analyze the dynamic behaviour of gears under rattle conditions. The DWT has been chosen because it is a particularly suitable instrument to recognize discontinuities, such as jumps or instantaneous changes in the slope of the signals, due to their localization in the time domain. Wavelet analysis allows to characterize an event by the localization, the kind, and the amplitude of the event. This technique enables, moreover, to define new indices for metrics of gear rattle especially useful in order to conduct comparisons among different operative conditions. Some examples of application of the proposed technique are reported in the paper. The theoretical investigations regard comparative analysis with respect to the amplitude and to the frequency of speed fluctuations and to the gears lubrication regime.