Statistical analyses to contain the variability of heavy duty truck noise for good noise control

摘要

To reduce noise inside trucks, an optimized material treatment package is developed by iterative testing. Therefore success depends on the stability of the truck noise output over this process, which is sensitive to variations in ambient, load, engine and tire conditions, as well as fuel weight and cetane rating, most of which can be monitored. These variations may be exacerbated by the feedback sensitivity of the engine calibration program, particularly the exhaust treatment protocol.Using multivariate and principal component statistical analyses of these variations, while running on a dynamometer in a hemi-anechoic chamber, we have developed a process that: Determines test repeatability, so the acquired data consists of representative runs, from controlled sources of variation, and monitored sources that cannot be controlledo Gives insight into effective test sample sizes for making decisionso Highlights unusual results to optimize consistency between iterations Determines the most critical sources that drive, beyond a norm, a corresponding variation in the acoustic spectra Ultimately, quantifies the acoustic benefit of an optimized noise control package within confidence limits of the testing