Fatigue of asphalt binders - investigation of edge effect phenomenon and development of surrogate test

This research was started as a continuation of the National Cooperative Highway Research Program Project 9-10 (NCHRP report 459). The project introduced a new laboratory method, called the time sweep, for measuring asphalt binder's resistance to fatigue damage. The method was challenged by a number of experts in the field claiming there are artifacts to the proposed test procedure which could significantly limit its usefulness. This research was conducted in two phases to examine these claims and to propose new alternatives fro measuring binder fatigue. In the first phase, this research examined the artifacts' claim using 2 different approaches and found the claims to be unfounded by showing the proposed testing procedure are reliable. However, the procedure was found to be too complex for practical application therefore a second phase of the research was initiated.;The objective of the second phase was to develop an effective and practical surrogate binder fatigue test. Two surrogate tests were attempted in the second phase; the Direct Tension Test (DTT) and the Dynamic Shear Rheometer (DSR) stress sweep test. Devices for both test (DTT and DSR) are used commonly in practice. The DTT approach was found to be highly influenced by variability and sample preparation; therefore it cannot be used as a general surrogate test. Data analysis of the DSR stress sweep approach showed that the stress sweep failure parameters correlate strongly with binder fatigue failure parameters as measured by the NCHRP method. This finding was verified by the comparison of the stress sweep failure parameters with the field performance of the binders used in the Federal Highway Administration (FHWA) Accelerated Loading Facility (ALF). The successful validation suggests that there is a high potential for the stress sweep testing protocol in the DSR to be used as a simple and practical test to characterize asphalt binder resistance to fatigue damage. The correlations with a limited sample size were not perfect and will need further validation in the future.