A Comparison of Measured and Predicted Low Temperature Cracking Conditions
Amy L. Epps
Abstract
The addition of crumb-rubber, manufactured from scrap tires, to the binder in asphalt concrete pavements has been suggested to minimize low temperature cracking. The modified binder lends enhanced material properties to modified asphalt-aggregate mixtures, increasing their resistance to induced tensile stresses caused by thermal changes. A second benefit arises from the reuse of scrap tires, alleviating part of the solid waste disposal problem.
To evaluate both crumb-rubber modified (CRM) and unmodified mixture resistance to low temperature cracking, three approaches were used to measure or predict temperature and stress at fracture. These fracture conditions were then compared to examine the compatibility of the selected approaches. Mixture response in terms of fracture temperature and stress was directly measured in the Thermal Stress Restrained Specimen Test (TSRST), and material properties measured in the Indirect Tensile Test (IDT) were used to predict these fracture conditions. Mixture fracture temperatures were also predicted from results of binder tests in the Bending Beam Rheometer (BBR).
TSRST results for one of the binders used in this study indicated that modified mixtures perform better in colder climates than their unmodified counterparts. In comparison, predicted fracture temperatures from BBR binder tests more closely agreed with predicted fracture temperatures from IDT material properties than with measured fracture temperatures from the TSRST. As a result, this analysis suggests that binder testing can provide predicted fracture temperatures in place of a complex analysis of expensive and time consuming IDT mixture tests. In general, TSRST fracture temperatures were more conservative (higher) than predicted fracture temperatures (IDT) if the two methods did not agree. In addition, mixture fracture temperatures predicted from BBR data did not highlight the improved performance of modified mixtures shown in TSRST results for one of the binders used in this testing program. Equivalent predictions for unmodified and modified mixtures emphasize that mixture testing may be necessary to characterize modified mixture behavior at low temperatures.
The results from this laboratory testing program indicate that there is strong agreement between the three methods for some mixture/cooling rate combinations, and only a few combinations illustrated large differences between measured and predicted fracture temperature and fracture stress. None of the methods was selected as the best or most accurate method, as the best method should match field performance in terms of resistance to low temperature cracking and field data were not considered as part of this study.