Abstract
In this study, the fatigue performance of 12 plant-produced mixtures from New Hampshire and Vermont that contain reclaimed asphalt pavement (RAP) contents of 0-40% by total weight of mixture was evaluated. The mixture tests included dynamic modulus, uniaxial fatigue, beam fatigue, and overlay tests. Also, the simplified viscoelastic continuum damage (S-VECD) model failure criterion, called the G
R
method, was applied and input to the linear viscoelastic pavement analysis for critical distresses (LVECD) programme to predict the fatigue behaviour of the tested mixtures on thin and thick asphalt pavement structures. In order to explain the observed fatigue behaviour, the performance grades (PGs) of the binders that were extracted and recovered from the mixtures were determined. In general, the addition of RAP resulted in an increase in the stiffness of the materials. The magnitude of the impact of higher RAP percentages varied with each set of mixtures. The S-VECD model and beam fatigue test data showed a loss of fatigue resistance for high-percentage RAP mixtures in most of the cases. The overlay tester results showed clear drops in performance at higher RAP contents. The impact of lowering the PG of the virgin binder to compensate for higher levels of RAP also was studied. Lowering the PG led to improvement in the fatigue properties and was found to be a convenient practice. The changes in the measured properties also appeared to be a function of mix design variables that included the stiffness of the RAP, asphalt content, and production parameters such as silo storage times. In some cases, the effects of these factors outweighed the impact of the RAP level or PG of the virgin binder in the mixtures.