Probabilistic Method of Asphalt Binder Selection Based on Pavement Temperature

Abstract

The Strategic Highway Research Program, conducted in the United States and Canada from 1987 to 1992, included a significant asphaltic concrete study. The study included a new grading method for asphalt binders, performance grading, resulting in what is termed PG asphalt. The grade selection process is based on air temperature distribution. Since 1994, the Long-Term Pavement Performance program has established a seasonal monitoring program that consists of a number of sites with air and pavement temperature sensors and a data logger that records the air and pavement temperatures hourly. These data are now available for prediction of the extreme temperatures that are needed for the selection of the appropriate PG asphalt binder. On the basis of the review of the recent pavement temperature studies and statistical analysis of the pavement temperature data available, an empirical prediction model has been developed from the original data set and validated from the expanded data set. The model is compared with existing prediction relationships including the heat flow model. Pavement prediction data and climatic data on high and low temperature from weather stations are used to calculate the probability that a specific asphalt temperature will be exceeded. Variations both in prediction and in annual maximum or minimum air temperature are assumed to be independent and considered in the calculation. The reliability contours for a given PG grade are presented as a convenient tool to the asphalt binder selection.