Morphological characterization and mechanical analysis for coarse aggregate skeleton of asphalt mixture based on discrete-element modeling

- Morphological characteristics of coarse aggregates were quantified.
- Morphology of coarse aggregates were reconstructed within PFC2D.
- Virtual penetration test was built to predict mechanical property of aggregates.
- Stronger aggregate skeleton leads to better penetration resistance.

Focused on the aggregate microstructures to characterize the aggregate mechanical properties, this paper proposed a modeling method of coarse aggregate based on the particle morphological characteristics. By using the Aggregate Image Measurement System (AIMS), the realistic shape of granular aggregates was captured and the statistical analysis was made to quantify the morphological differences firstly. Then by using Discrete-Element Method (DEM) software named as Particle Flow Code in two dimensions (PFC2D) and with the help of the image processing techniques, an algorithm was developed to model the two-dimensional shape of aggregates. Through linking the shape to the morphological index for each aggregate, the coarse aggregate skeleton was built which is consistent with the actual composition and structure of the laboratory specimens. After that, based on laboratory penetration test for granular aggregates, virtual penetration test was built by PFC2D to verify the validity of the developed modeling algorithm. An optimized method was proposed further then to improve the simulation efficiency. This is because of the time-consuming process when modeling the actual specimen which was composed of numerous coarse aggregates with various shapes. Some representative particles were selected and reconstructed based on the angularity index to form the virtual specimens with aggregates of gradations AC-13 and SMA-13. At last the virtual penetration test of AC-13, SMA-13 was conducted to predict the mechanical behavior and the coarse aggregate skeleton structure characteristics were analyzed from the point of micro-contact state. It is proved that the proposed modeling algorithm could well capture the realistic shape of aggregates and virtual test based on the aggregates model could characterize the granular aggregates mechanical properties accurately.