Search

An ultrathin seal serves as a preventive maintenance technology. The design method predominantly relies on engineering experience and lacks a theoretical foundation. In order to investigate the impact of aggregate size, spreading mass, and original pavement texture depth on the formation of an ultrathin seal, a compaction model was developed using the discrete element method (DEM). Different aggregate sizes and spreading masses were analyzed to examine the particle structure and thickness. The three-dimensional (3D) Weierstrass–Mandelbrot (WM) function was used to reconstruct random rough surfaces to assess the effect of the original pavement texture depth (1, 2, 3, and 4 mm) on voids and dosage of filling asphalt. Results showed that the dynamic compaction led to a rapid increase in collision velocity and then declined sharply by 0.02 s, eventually reaching zero by 0.3 s. Particle contacts steadily increased to 365 by 0.3 s and kept stable after the plateau, indicating the formation of an interlocking structure. Coordination numbers and contact vectors varied with particle size and spreading mass, which were more pronounced for small particles. However, the smaller particles (<4.5 mm) had negligible compaction effects, suggesting that the optimal aggregate size should be 3–5 mm. Excessive spreading mass resulted in a multilayered structure incompatible with the single thickness requirement and the recommended range was 9–12 kg/m2. Furthermore, greater texture depth led to higher void content at the bottom of the seal, requiring more filling asphalt. Therefore, when designing asphalt dosage, both the compaction and the variation in voids due to the original pavement texture depth should be considered to ensure sufficient contacts. [ABSTRACT FROM AUTHOR]

To further facilitate the automation and computational efficiency in asphalt mixture micromechanical simulation using finiteelement methods, this paper develops a three-dimensional (3D) microstructural modeling method for asphalt mixture to automatically generate the adaptive 3D finite-element mesh. The methodology contains three main steps, as follows: (1) two-dimensional (2D) enclosed outlines in every X-ray computed tomography (CT) image of the specimens are detected with shape features of outlines and grayscale discrepancy between each two different phase pixels; (2) for each independent geometric solid such as aggregate particle and air void in a specimen, the mapping between the solid and its outlines on various layers is established, so that the 3D model of each solid is constructed; and (3) models of the same phase are assembled into a specific substructure model. The specimen microstructure model is finally obtained with appropriate Boolean operations towards specific substructure models. Experiment results indicate that the specimen microstructure model not only has an excellent accuracy in modeling and simulation, but also supports adaptive meshing toward each specific substructure model, which significantly reduces the computational cost and can be applied to asphalt mixture numerical simulation. [ABSTRACT FROM AUTHOR]