Molecular simulations of asphalt rheology: Application of time-temperature superposition principle

We show that the several decades gap between the time scales of experiments and atomistic simulations can be significantly reduced by applying time-temperature superposition (TTS) to the simulation results for various rheological properties of asphalt. Molecular dynamics simulations with atomistically detailed models were performed to characterize the temperature dependence of shear viscosity, dynamic modulus, and tensile creep compliance of AAM-1 asphalt. The TTS principle was successfully applied to the data to construct master curves of these rheological properties, and a comparison of simulation data with experiments showed good agreement. Application of TTS resulted in a six decade increase in the frequency range available for moduli determination and a two decade increase in both the shear rate range available for viscosity determination and the time scale available for characterization of creep compliance. The horizontal shift factors (a_T) used in the construction of the master curves for the different rheological properties were quantitatively consistent with each other. The corresponding temperature dependence of the scaled relaxation times was compared with the expectations from three literature models, namely, the Vogel-Fulcher-Tammann, double exponential, and parabolic-Arrhenius models. A distinct change from super-Arrhenius to Arrhenius behavior was observed for the scaled relaxation times below the glass transition temperature (T_g). Only the parabolic-Arrhenius model was able to capture the temperature dependence of the data over the entire temperature range that covers both the glassy and the rubbery states.