Performance of Indiana’s Superpave HMA Mixtures

This project followed the performance of a number of hot mix asphalt paving projects for seven to eight years in an attempt to relate laboratory test results to field performance. First, a small scale study compared sampling locations, specific gravity test procedures and compaction devices used on two samples from projects, one with volumetric acceptance and one with non-volumetric acceptance, to determine the best combination to use for Part 2 of the study. In Part 2, 12 projects constructed in 2001 and 2002 were sampled, tested and monitored. Laboratory testing involved determining maximum and bulk specific gravities, binder content, air voids and other properties of the mixes. Laboratory performance tests were also conducted, including Superpave shear tests at high and intermediate temperatures and indirect tensile tests at low temperatures. The Purwheel loaded wheel tester was used to test mixtures from four of the projects to examine the mixes’ tendencies to strip and rut. The laboratory testing results generally predicted that all of the mixtures would perform well in the field in terms of rutting. Results ranged from fair to excellent, with the vast majority of the results in the good to excellent categories. The indirect tensile testing results did not indicate any extreme problems would be expected with thermal cracking of these materials. While some of the mixes did have expected critical cracking temperatures warmer than the binder low temperature grade, all of them would be expected to perform well down to about -17°C. The single mix with a -28 grade binder, as opposed to the -22 grades for all of the other mixes, also had the lowest critical cracking temperature. Field performance of these projects was monitored through a condition survey and using videologs, rut depths and roughness measurements from the INDOT Pavement Management System. Through 2009, all 12 projects were performing well with rut depths generally less than 0.1 inch (2.5 mm) and roughness of less than about 100 in/mi. While there is some cracking on many of the projects, it is generally not severe and is likely reflective rather than thermal cracking. In general then, the laboratory results agree with the field performance. The lab tests suggested that the mixes would be resistant to rutting and fairly resistant to thermal cracking, and this was verified by the field performance in all cases. This study is limited somewhat by the fact that none of the mixes would be expected to perform poorly. Having mixes that “failed” a test would help to establish the boundaries between good and bad performance. The study is also subject to a common limitation of long-term performance studies – the fact that technology has evolved over the course of the project. Mix design parameters have changed somewhat and new test methods, particularly the dynamic modulus and flow number tests, have gained prominence. Nonetheless, this study does show that Superpave mixes from the early part of the century can be expected to perform well and that the laboratory performance tests used in this study generally can predict this good performance