Performance Predictions of Rutting for the National Center for Asphalt Technology Test Track

Performance-related specifications base pay rates on tests that measure and predict the performance of the mixture. This study evaluated several methods for predicting the rutting resistance of asphalt mixtures. The mixtures and field performance data came from the National Center for Asphalt Technology test track. The pavement sections evaluated were part of an experiment that focused on warm-mix asphalt, high-content reclaimed asphalt pavement (50%), and a combination of the two. One method involved using the dynamic moduli of the mixtures in confined and unconfined conditions. The ranking of the mixtures predicted from both conditions was good compared with the observed ranking from the track but was more reasonable with the confined data. These predictions were compared with common current tests such as the asphalt pavement analyzer and the Hamburg wheel tracker tests. From the ranking of rut depths, these tests did not appear able to distinguish subtler changes to material properties needed for a performance-related specification. Another method employed a viscoplastic shift model implemented within the layered viscoelastic continuum damage program to predict the rut depth. The predicted rut depths were biased 2.5 mm higher than those measured in the field when all three layers (175 mm) of asphalt were used to predict the rut depth, but the ranking matched the field quite well. The results from the top two layers (100 mm) of asphalt were less biased but were more variable than the three-layer results. The results suggest that all pavement layers should be evaluated for predicting rut depths, not just the top two layers.