Nonlinear modelling of selected micro- and macro-properties of weathered asphalt mixtures

Asphalt mixture specimens with three different gradations of AC-13, SMA-13, and Superpave-13 were aged in an accelerated weathering machine (AWM) for a duration of 0, 1000, 2000 and 3000 h. Gel Permeation Chromatography (GPC), Fourier Transform Infrared Spectroscopy (FTIR) and Brookfield viscometer tests were performed on the recovered asphalt binder samples from the weathered mixtures. Nonlinear models fitted the measured properties by a method of LM+ universal global optimization of 1stOpt software. The results showed that: 1) Nonlinear models fitted the measured properties well, the rate and the degree of aging of the asphalt binders can be well explained by the parameters of these models. 2) As the aging duration is increased, Brookfield viscosity of the weathered asphalt binders increased gradually, and the percentage of Large Molecule Size (LMS) from GPC and the functional group of carbonyls (C=O) from FTIR increased as well. 3) The calculated values by the established models were highly correlated with the measured values of the properties of the weathered asphalts.