Single vehicles’ noise emission curves analysis by means of first and second derivatives

In the last decades, car manufacturers have gathered efforts to lower vehicle noise emissions by acting on power-unit insulation and designing advanced mufflers. Moreover, the penetration rate of electric vehicles in the national fleet composition is increasing. As a result, the rolling noise contribution, which historically overclasses the other vehicle noise contributions starting from 40 to 50 km/h, could begin being relevant at lower speeds. Detecting the contribution of each sub-source to the overall vehicle can be challenging and costly due to the need for extensive measurement campaigns. Thus, this paper aims to investigate the aforesaid hypothesis through the analysis of the first and second derivatives of vehicles’ noise emission curves which does not require broad data collection. Noise emission curves for seven probe vehicles, differently powered (namely, two diesel, two gasoline, two hybrid-electric, and one LPG), were obtained on flat road locations, and, for three of them, on a hilly road location. Subsequentially, the first and second derivatives of the noise emission curves were computed through the finite difference approach, to identify the speed regions where rolling noise contribution starts becoming predominant. Results showed that the rolling noise contribution is relevant for speeds lower than 40 km/h at flat roads, and it moves forward on the hilly road location. The analysis also revealed that differences in terms of noise emissions between hybrid and conventional vehicles are detectable in the speed range between 15 and 20 km/h, where the former category of vehicles could work purely in electric mode