Beyond the PG specification for asphalt binders

Classification of asphalt binder has evolved since the invention of asphalt pavement in 1870. Starting with penetration grading, moving forward to viscosity grading, and now the current system of performance grading (PG). All states in the US now adhere to some form of the PG system. During the last decade, several new modifiers and extenders have been introduced to modify the grade of a straight-run binder to achieve a target PG. In some cases, it has been observed that although a binder may meet the current PG specification, it may result in significantly sub-par performance. This suggests that the current PG specification does not accurately capture the performance characteristics of a binder. The main objective of this study was to use alternative chemical and mechanical tests on a large set of binders to identify differences in binders with a similar PG. This study examined the performance of 34 asphalt binders from 12 different binder sources using standard PG tests following AASHTO M320, as well as tests beyond the PG specification. The tests outside of the PG framework included chemical tests (X-ray fluorescence and spot), as well as other mechanical tests (Multiple Stress and Creep Recovery to measure permanent deformation potential at multiple high temperatures, BBR Pro to measure tensile strength at low temperatures, and poker chip test to measure tensile strength at intermediate temperature). For tests outside the PG specification, outlier criteria were developed based on the results found. Outliers were defined differently for each test and do not represent positive or negative outcomes in terms of expected performance. The tests outside of the PG specification (spot, ΔT [subscript c], aging sensitivity, poker chip, and low temperature strength) all produced many outliers, some even extreme outliers for binders with a similar PG. This study highlights the need for additional testing beyond the PG specification to improve binder grading. More performance testing on mixtures is needed to establish if outlier behavior is beneficial or detrimental.