New Test Protocol to Measure Fatigue Damage in Asphalt Mixtures

ABSTRACT The measurement, analysis and interpretation of hot-mix asphalt concrete fatigue tests have been a subject of research for many years. Standard flexural tests using ectangular or trapezoidal beams are non-homogeneous in that the stress-strain fields and the temperature distribution within the test specimens are unknown, making a rigorous analysis of the phenomena that occur during these tests very difficult. A new protocol for fatigue testing is proposed using a uniaxial “push-pull” test with three loading stages on a cylindrical specimen. The first and last stages are at very low loading levels, below the endurance limit of the mixture being tested. Fatigue damage occurs during the second stage of loading when the applied stress or strain is sufficient to cause fatigue damage. The difference between the moduli at the end of the first and at the end of the third stage of loading is caused by fatigue alone. The presence of an endurance limit is verified and it is shown that the drop in modulus during a standard fatigue test is primarily caused by nonfatigue phenomena. The rate of fatigue damage obtained with the proposed protocol is approximately one tenth of that predicted by standard tests. The effects of non-fatigue phenomena such as self heating, self cooling, and nonlinearity are discussed, as well as a phenomenon that the authors, for want of a better term, refer to as thixotropy. The authors conclude that the major portion of the modulus lost during a fatigue test, as well as modulus gained during periods of rest or decrease of loading, is the result of thixotropy.