Modification of Dynamic Modulus Predictive Models for Asphalt Mixtures Containing Recycled Asphalt Shingles

Transportation agencies are increasingly using recycled asphalt shingles (RAS) in asphalt pavements for economic and environmental benefits. The primary components of RAS are asphalt, mineral filler, aggregate granules, and either organic or fiberglass fibers. The effect of these components on the dynamic modulus of an asphalt mixture, which is a key input in the Mechanistic-Empirical Pavement Design Guide (MEPDG) and a critical property of asphalt mixtures related to its performance, are still uncertain. As part of National Pooled Fund Study TPF-5(213), a series of experimental plans was conducted utilizing demonstration based paving projects to investigate various factors related to RAS utilization. Thirteen mix designs with RAS contents ranging from zero to six percent were developed and constructed in Indiana, Iowa, Minnesota, and Missouri. Field produced mixtures were procured and sent to Iowa State University’s Asphalt Lab for laboratory dynamic modulus testing. The test results are used in this study to evaluate two commonly used dynamic modulus predictive models, the Witczak and Hirsch models. Two versions of the Witczak models, which were developed in 1999 and 2006, are evaluated in this research. It was found that the Witczak models were not very effective in estimating the modulus values of RAS mixtures. This study concludes that commonly used dosages of RAS in asphalt mixtures affects the prediction accuracies of the Witczak and Hirsch models due to biases of the E* predictive models in terms of input parameters, inaccuracies of input parameter measurements, effects of blend efficiency, and influences of the RAS fibers.