Test pavements were constructed with the primary variable being the binder type. Full-scale rutting and cracking performance were collected and compared to various candidate binder tests to identify the most discriminating element to replace the current specification for asphalt fatigue cracking and rutting.
The primary objective of this full-scale accelerated pavement testing was to evaluate the performance of unmodified and polymer-modified asphalt binders, and to recommend improved specification tests over existing Superpave (SUperior PERforming Asphalt PAVEments) performance-grading methodologies.
Two fatigue cracking specification tests were identified as being more capable than others: binder yield energy and critical tip- opening displacement. Two rutting specification tests that quantify irrecoverable deformations exhibited the best strength to capture rutting: multiple stress creep, and recovery and oscillatory-based nonrecoverable stiffness. Based on the full-scale performance and laboratory tests, crumb rubber (recycled tires) modified asphalt (Arizona wet process) was shown to significantly slow or stop the growth of fatigue cracks in a composite asphalt pavementstructure. A hybrid technique to modify asphalt with a combination of crumb rubber and conventional polymers (terminally blended) exhibited good fatigue cracking resistance relative to the control binder without the special handling procedures needed for some crumb rubber modified asphalts. Also, a simple addition of polyester fibers-to-asphalt mix was shown to have high resistance to fatigue cracking without the use of polymer modification. The research study also quantified the capabilities of the National Cooperative Highway Research Program’s mechanistic-empirical pavement design and analysis methodologies to predict rutting and fatigue cracking of modified asphalts that were not captured in the calibration data from the Long-Term Pavement Performance program. Falling weight deflectometer, multidepth deflectometer, and strain gauge instrumentation were used to measure pavement response. The results illustrated that the globally calibrated mechanistic-empirical performance models could differentiate between structural asphalt thickness but had difficulty differentiating modified from unmodified asphalt binder performance. Nonetheless, the mechanistic-empirical performance ranking and predictions were enhanced and improved using mixture-specific performance tests currently being implemented using the Asphalt Mix Performance Tester.