Use of image analysis to predict the sound absorption coefficient of bituminous mixtures

Sound absorption coefficients of bituminous mixtures are predicted owing to image analysis. A three-parameter semi-phenomenological model, well adapted to porous road materials, is used. Three geometrical parameters: porosity, resistivity and tortuosity, are estimated. Image analysis on high-quality digital photographs allows one to obtain porosity and pore-size distribution. A specific relation is assumed to express tortuosity in function of porosity. This relation has been already used in previous papers for granular media and porous asphalts. It is shown that the obtained pore-size distribution is close to a log-normal distribution so by using log-normal distribution properties, resistivity is then determined. The predicted parameters are compared to direct measurements. The proposed methodology is tested on three bituminous mixtures commonly used for quiet roads.