Effect of Property Variations of Raw Materials on the Variability of Dynamic Moduli of Asphalt Mixtures

摘要

This paper presents a research on the variability of dynamic modulus of asphalt mixtures because of the variations of raw materials by error propagation theory. The coefficient of variation of dynamic module, COVE, was derived from the η-based predictive model and the variance propagation model. Then, the raw materials' properties related to the dynamic modulus of asphalt mixture were measured and collected, and the coefficients of variation (COV) were finalized and substituted into the COVE expression. The effects of variations of raw materials' properties on the variability of dynamic module were investigated, and the efficiencies and relative efficiencies of reducing materials variability on decreasing COVE were also calculated. It is found that V_a makes the greatest contributions, followed by ρ_(4.75) and V_(beff). At high temperature the contribution of viscosity is similar to that of V_(beff). but is negligible at low temperature. However, the contributions of others are very small in all conditions. Reducing COV of ρ_(4.75) is with the most efficiency to decrease COVE and reducing COV of V_a is with the most relative efficiency. Efficiency of reducing COV of binder viscosity varies greatly with the temperature and frequency, whereas reducing COV of other parameters are much less efficient.