Modeling the Tire-Pavement Noise as a Regression Function of Speed, Mixture Properties, and Age of the Layer

This paper focuses on the development of a statistical model that estimates the close proximity level (CPX-L) of bituminous mixtures usable in urban, extraurban, and motorway road networks as a function of survey speed, aggregate gradation, bulk properties, and age of the layer. The proposed equation derives from an existing model developed on Italian urban and extraurban roads whose coefficients were recalibrated through analysis of variance by using data from mixtures used in the Italian motorway network. The original equation was further refined by introducing the age of the layer as a new covariate to predict the durability of the acoustic performance. The age of the layer resulted in a higher statistical relevance than the cumulative traffic for predicting acoustic degradation. A unique aging coefficient of $\sim 0.03\pm 0.01\mathrm{dB}(\mathrm{A})/\mathrm{month}$ was calculated for dense, porous, and semiporous mixtures installed in similar Italian intermediate-mild climatic regions. This value predicts an increase in the tire-pavement noise of $\sim +0.4\mathrm{dB}(\mathrm{A})/\mathrm{year}$ and $\sim +2.2\mathrm{dB}(\mathrm{A})$ in 5 years, which is compatible with other research projects and European noise regulations. The effects of the mixture type were also investigated, providing a negligible variability of $+0.3\mathrm{dB}(\mathrm{A})/5\mathrm{years}$ if only porous and semiporous mixtures are considered. The use of the model is recommended between 50 and $90\mathrm{km}/\mathrm{h}$, where the current equation explicates almost 94% of the variability and the predicted values show an accuracy compatible with the precision of the CPX method. Future works can contribute to further improving the accuracy and the range of applicability of the model providing also more detailed insights into the acoustic aging of different mixtures.