Analysis of Gear Rattle Noise and Vibration Characteristics Using Relative Approaches 2016-01-1121

Noise signals of the driver’s right ear include those of engine, environment, chassis dynamometer, loaded gears and unloaded gears when they are recorded in full vehicle on chassis dynamometer in semi-anechoic room. Gear rattle noise signals of the driver’s right ear caused by unloaded gear pairs can’t be identified or quantified directly. To solve the problems, relative approaches are used to identify and quantify the gear rattle noise signals. Firstly, the rattle noise signals of the driver’s right ear are filtered by human ear characteristic functions and steady noise signals are extracted by regression and smoothing processes. The noise signals are regressed at 200ms interval in the hearing critical frequency bands and smoothed in the flanking frequencies. Then, the noise relative approaches are obtained by subtracting the steady noise signals from the filtered noise signals, which are the transient noise signals of the unloaded gear pairs inducing the rattle noise. Finally, the occurrence time, frequency and level of the rattle noise are obtained from the noise relative approaches. Through this method, ten rattle noise signals of the highest relative approaches are identified within 3.5 s in the full vehicle test. The results show that the occurrence time obtained by the rattle noise relative approaches is entirely coincided with that of the ten rattle noise signals which is recorded in the raw noise signals playback. The results of the rattle noise identification are consistent with that of the subjective perception. Thus the gear rattle noise is identified accurately. The interference of the noise signals of the engine, the chassis dynamometer, the environment and the loaded gears is effectively removed by the noise relative approaches to avoid false diagnosis. Gear rattle vibration signals on transmission housing obtained in the full vehicle test are similar to those of the rattle noise. Therefore, the occurrence time, frequency and level of the rattle vibration also can be obtained by the relative approaches.

Results show that bigger relative approach represents higher gear rattle noise and vibration level. The gear rattle vibration signals of the transmission housing always coincide with all the rattle noise signals of the driver’s right ear which delay the rattle vibration signals for 0.01∼0.04 s. Thus the transmission housing is the major transfer pathway of the gear rattle noise. The frequency range of the gear rattle noise is within that of the rattle vibration because of the noise transfer function from the transmission housing to the driver’s right ear. The gear rattle vibration characteristics of each surface and direction are different. Therefore, the rattle anisotropy can be considered when rattle mechanism is investigated. The rattle vibration of the upper, front and left surfaces on the transmission housing is higher than other surfaces. Those three surfaces are the major improvement targets for optimization of the rattle performance.