Research and Comparison of Nano-Asphalt Mixture Fracture Toughness Based on Machine Learning Technique

Low-temperature cracking (LTC) is a critical form of pavement distress in cold regions. The fracture toughness in the semicircular bending (SCB) test serves as an indicator of LTC growth. Firstly, this study evaluated the effect of adding nano Al2O3 on the improvement of hot mix asphalt (HMA) fracture toughness. Another goal of the paper was to investigate the influence of different parameters, such as temperature (-5, -15, and -25 °C), loading mode (I, II, and I/II), crack geometry (vertical and angular cracks), and nano-modification, on the fracture toughness of HMA by using machine learning technique. An artificial neural network (ANN) was employed to quantify the impact of these parameters. The findings of this research clearly show that although asphalt mixtures in cold region are prone to thermal cracks, the addition of nano Al2O3 improves their resistance by 12% in comparison with control mixtures. The ANN analysis identified loading mode is the most significant factor affecting fracture toughness (48% contribution). Temperature followed with a 28% contribution, while crack geometry and nano Al2O3 modification each contributed 12%.