Bayesian Method to Determine the Dynamic Material Characteristics of Hot-Mix Asphalt

Abstract

A reliable method for determining the dynamic material characteristics of hot-mix asphalt using a Bayesian method based on Latin hypercube sampling, impact resonance testing (IRT), and the shift factor of linear viscoelastic (LVE) asphalt concrete specimens is reported. Discrete resonance moduli data were obtained from the IRT at temperatures of 5, 25, 40, and 50°C. The shift factor of the LVE was used to translate the discrete points of resonance moduli to higher or lower frequencies, depending on the temperature of the specimen. Based on the temperature–frequency combinations, Bayesian statistical predictions were used to create a dynamic modulus master-curve representation, using the resonance moduli data and Latin hypercube sampling. The results for three different hot-mix asphalt mixtures were in good agreement with dynamic moduli data obtained using other testing methods.