Characterizing Properties of Asphalt Cement at Cold Temperatures

Abstract

A range of binder test results for Strategic Highway Research Program (SHRP) Materials Reference Library asphalts were used to identify critical low-temperature behavior. Test results were obtained from in-house testing, Minnesota Department of Transportation (Mn/DOT) testing, and previously reported SHRP research reports. Tests used were the Fraass brittle point, bending beam, a dynamic shear rheometer (DSR), and differential scanning calorimetry (DSC) glass transition temperature. A good correlation between the Fraass brittle point and the bending beam rheometer limiting stiffness indicated that the Fraass brittle point represented the behavior of a binder at a stiffness of approximately 200 MPa. Because the traditional method of using the DSR to obtain the glass transition temperature, Tg, from the peak of the loss modulus curve did not provide sufficiently specific results, the temperature at which a minimum phase angle was used to define the Tg instead. These results showed a good correlation with the bending beam results and a strong correlation with the Fraass brittle point. The DSC results were not well correlated with either the rheometer results or the Fraass brittle point. This appears to be due to the intricacies related to interpreting DSC data. Binder results were compared with results reported by SHRP researchers for the indirect tensile creep and thermal stress restrained specimen test. Results indicated that cold-temperature mixture properties are predominantly due to binder properties. Results also indicated that the asphalt source was more important than the asphalt grade in defining cold-temperature properties. The two binders used to construct the Minnesota Road Research facility on I-94 in Otsego were classified by Mn/DOT as PG 58-28 and PG 58-22, respectively. The Canadian SHRP researchers’ equation for estimating the minimum design pavement temperature indicated a cold-temperature design temperature of —28°C, but the SHRP 90 percent reliability grading indicated that a PG 58-40 was needed. Evaluation of the pavement performance over two winters—the first of which was extremely cold even by Minnesota standards—showed that the PG 58-28 had no thermal cracks while the PG 58-22 test sections had 12 transverse cracks. This indicates that the Canadian SHRP equation may be considered a high-reliability minimum design temperature.