Predicting rutting performance of asphalt mixture from binder properties and mixture design variables

Chuanqi Yan,Yuan Zhang,Hussain U. Bahia

DOI: https://doi.org/10.1080/14680629.2020.1820890

2020-09-18

Road Materials and Pavement Design

Abstract:Rutting is one of the main asphalt pavement distresses, and it is believed that binder properties play an important role in mixture's potential rutting resistance. Generally, in the Superpave design system asphalt binder is selected based on the binder PG grade required for the climate. However, to implement performance control through the traditional Superpave PG system during mix design faces more challenges due to the changes of oil refinery, asphalt binder modification, and the use of recycled materials in the mixture. This paper aims to investigate the role of asphalt binders and mix designs in the rutting resistance of asphalt mixtures. A total of 22 plant-produced mixtures from 10 states and 5 laboratory-produced mixtures were investigated for their rutting performance in this study. Those tested mixtures include various binder grades, aggregate sources, RAP contents, and mix designs. The Superpave binder rutting parameter G*/sinδ and Multiple Stress Creep and Recovery (MSCR) grade parameter Jnr3.2 are correlated with the mixture dynamic modulus and Hamburg Wheel Tracking (HWT) rutting parameters. In addition, the effects of other mix design variables on the rutting resistance are evaluated using a statistical analysis. The results show that by incorporating the binder properties and mix design variables, the stepwise regression fitting provides a much better prediction for HWT rutting parameters as compared to using only the binder properties. This indicates the importance of proper selection of the binder grade for mixture performance, and it also calls for justifying more focus on including mixture design variables to ensure acceptable mixture rutting performance. The statistical analysis can provide an insight into the use of artificial intelligence and machine learning by generating mathematical models for evaluating which mixture design variables are needed for better prediction of rutting performance.

construction & building technology,materials science, multidisciplinary,engineering, civil