Predicting Asphalt Superpave Properties

摘要

Superpave performance grading of asphalt addresses the limitations of empirical asphalt physical property tests such as penetration, viscosity, and ductility. However, it is possible to accurately predict the Superpave grade with a reduced set of testing under certain conditions. This paper describes a methodology that uses three measurements on original asphalt: penetration at 25℃, kinematic viscosity at 135℃, and Dynamic Shear Rheometer measurement on Original asphalt (DSRo G*/sinδ) to predict low-temperature properties and determine Superpave grade. The high-temperature specification is determined from measuring DSRo G*/sinδ. Multiple linear regression was used to develop predictive equations for Bending Beam Rheometer (BBR) stiffness and Slope of the Creep Curve (m-value). The equations allow for the determination of the low-temperature specification temperature. Superpave tests themselves have high single operator and multi-lab variability due to the nature of the tests. The multi-laboratory precision of m-value by American Association of State Highway and Transportation Officials (AASHTO T-313) is equivalent to almost one-half the Superpave grade to grade range. A more precise determination of m-value is possible using predictive equations. Regression analysis indicates that the predictions for m-value and stiffness are within the precision of the actual Superpave tests for most of the asphalt samples studied.