Aggregate packing is a key factor controlling the durability of asphalt mixtures. Considering the complex system of the aggregate structures, multiscale method was utilized by dividing asphalt mixtures into four scales, including asphalt binder, mastic, mortar and asphalt mixtures. Six asphalt mixtures with different gradation types and nominal maximum aggregate size (NMAS) were prepared and the corresponding mastics or mortars were also designed. Comprehensive image analysis methods, including high-resolution scanning, Scanning Electron Microscope (SEM), and Laser Diffraction Analyzer (LDA), were utilized to obtain the 2D images of asphalt mixtures, mortars and mastics. Then, the distribution curves of the binder within mastics, mastic within mortars, and mortar within asphalt mixtures were calculated and compared. New packing indexes were proposed, such as: expected value (Ebinder) and peak value (Pbinder) of the binder distribution at mastic scale, expected value (Emastic) and peak value (Pmastic) of the mastic distribution at mortar scale, expected value (Emortar) and peak value (Pmortar) of the mortar distribution at mixture scale. By comparing with volumetric parameters, it was found that Ebinder and Pbinder could correlate well with the volume ratio of filler to mastic. Emastic could correlate with the volume fraction of fine aggregate within the mortar, but the Pmastic may also influence by the fine aggregate gradation. The void filled with asphalt (VFA) has weak correlation with both packing indexes, Emortar and Pmortar, which indicates that VFA has limitation for evaluation of the packing properties of asphalt mixtures, and more indexes from the internal structure need to be considered.
Read full abstract