Probabilistic modeling of fatigue damage in asphalt mixture

Abstract

Fatigue cracking is considered as one of the major distress in asphalt pavements. Environmental factors like aging and moisture intrusion further complicate the cracking process. In this study, the effect of two levels of aging and three levels of moisture conditioning on the fatigue damage of asphalt mixtures was determined using a probabilistic approach. The fracture energy, the stiffness modulus, and the fatigue life of four different asphalt mixtures were determined at 15 °C and 25 °C. It was noticed that the stiffness modulus and fatigue life of asphalt mixtures increased with an increase in the level of aging and reduced with an increase in moisture conditioning cycles. At three freeze–thaw cycles, the reduction in fatigue life of asphalt mixtures was much higher compared to the stiffness modulus of the asphalt mixtures. It was noticed that moisture had a higher negative impact on the fatigue life of asphalt mixtures at 15 °C as compared to 25 °C. Both stiffness modulus and fatigue life of asphalt mixtures reduced with the increase in test temperature. The fatigue damage of unconditioned and moisture conditioned asphalt mixtures followed a lognormal distribution. The lognormal distribution parameters were used to determine the probability of fatigue damage at each loading cycle. It was noticed that a moisture conditioned asphalt mixture had a higher probability of fatigue damage compared to unconditioned asphalt mixture at different aging condition and test temperature. Asphalt mixtures subjected to three freeze thaw cycles had a higher probability of fatigue damage compared to one freeze thaw cycle. Further, the probability of fatigue damage was higher in the case of long term aged asphalt mixtures when compared to short term aged asphalt mixtures at a lower temperature of 15 °C. The damage based fatigue life prediction model had a good coefficient of determination for different conditions evaluated in this study. A Monte Carlo simulation with one million iterations was used to simulate the fatigue life of the damage based fatigue life prediction model. It was observed that a moisture susceptible asphalt mixture might not show an increase in fatigue life with an increase in the aging level.