Wind Noise Contribution Analysis

This article is motivated by observations of the wind tunnel measurement data acquired during benchmarking and program development for a variety of passenger vehicles over the years. In wind noise development, contribution analysis is a common practice to screen and identify the most significant sources and paths. In order to shed light on the whole picture of the contribution analysis, the work presented in this article falls into two categories. One is the analysis of underlying mechanisms for a better understanding of the phenomena observed in the contribution results. The other is the summarization of wind noise contributions obtained by wind tunnel testing for some representative subsystems, e.g., the contributions based on different reference states, the effect of grilles, underbody, acoustic glass, and auditory masking. A close look at the obtained numbers for each vehicle reveals that all these numbers have their intrinsic characteristics, and the same number may not tell the same story. The components with the same design, cost, and quality behave differently in vehicles with different wind noise levels. This work shows that contributions are generally reference-based and vehicle-dependent, and vary when the reference state is deviated even for the same vehicle. The same subsystem contributes more to a quiet vehicle than a mediocre vehicle, thereby a direct comparison of contributions among different vehicles is unfair and biased against quieter vehicles. As a first approximation, a quantitative estimation is derived to promote a qualitative understanding. It is used to facilitate a fair comparison of contributions, which are based on different reference states from either the same vehicle or different vehicles. It also implies that a vehicle with superior wind noise performance carries with it a much more stringent standard in subsystem design than an average vehicle even though their contribution targets are similar. The understanding of the work presented in this article would further benefit the interpretation of various contribution results, their comparison, and subsystem target setting.