Diffusion-Based Nanostructure and Self-Healing Characterization of Asphalt Binders Using Molecular Dynamics Simulations

Abstract Conventional macroscale and microscale experiments have not fully elucidated the thermodynamical behaviors of asphalt binders’ self-healing, particularly under varying thermal conditions and fracture dimensions, limiting understanding of nanoscale molecular processes. This research quantifies density changes and healing rates during self-healing, examining how thermal conditions and fracture dimensions influence asphalt molecules behaviors through molecular dynamics (MD) simulations. Asphalt binder models were developed using the four-component saturate, aromatic, resin, and asphaltene (SARA) system. Simulations were performed using Large-Scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) under different thermal conditions (30°C, 60°C, 90°C, and 120°C). Diffusion coefficients were determined through the Fick’s second law. To assess effects of damage severity, models with varying nanoscale fracture widths were created by introducing vacuum layers (5, 8, and 10 Å) between intact asphalt molecular layers. The findings revealed that self-healing occurs in three stages: (1) molecular contact, where the crack-tip molecules come into contact, (2) density restoration, where crack region density is restored, and (3) complete recovery, where the microcrack is fully healed. The self-healing rate peaks at the molecular contact stage, accompanied by a significant increase in density. It then approaches zero during the complete recovery stage, as the crack region’s density becomes consistent with that of the surrounding asphalt. Both reduced crack width and increased temperature promote molecular contact, enhancing the self-healing rate, with crack width has a greater influence than temperature. This MD-based investigation offers essential insights into asphalt binders’ self-healing mechanisms, facilitating the advancement of innovative additives and methodologies to enhance healing effectiveness. This study is the first to quantify the diffusion coefficient variation curve of three-stage self-healing (from 7.778 × 10 − 3 to 0.0323 × 10 − 3 m 2 / s ) through cross-simulation of crack widths ranging from 5 to 10 Å and temperatures from 30°C to 120°C. quantifying the diffusion coefficient variation curve (from 7.778 × 10 − 3 to 0.0323 × 10 − 3 m 2 / s ) during three-stage self-healing. It also confirms that crack width exhibits higher sensitivity to self-healing rate than temperature (with enhancement differences exceeding 50%).