Spray polyurea elastomer (SPUA) modifiers are finding increasingly widespread application in road engineering, particularly due to their significant importance in enhancing pavement performance in terms of high-temperature rutting resistance and fatigue resistance. To gain an in-depth understanding of the physicochemical properties of SPUA powder as a modifier, especially the damage mechanisms of its key functional spherical microstructures during the fine grinding process, this study prepared SPUA powder materials with different particle sizes and developed a novel quantitative method to assess the extent of their typical microstructural damage. Multiscale characterization combining x-ray diffraction, Fourier transform infrared spectroscopy, and thermal analysis (DSC/TGA) confirmed the material’s amorphous structure, intact functional groups, and thermal stability. Laser particle size analysis with scanning electron microscopy revealed surface-located spherical microstructures (50-200 μm). A ResNet18 convolutional neural network was developed to quantify structural damage in SEM micrographs through automated feature recognition. Results showed that progressive particle size reduction from 0.3 to 0.075 mm induced exponential microstructure degradation, with pore damage rates increasing from 21-37 to 61-72% (p < 0.01). The machine learning framework established quantitative correlations between grinding intensity and functional structure preservation, demonstrating 92.7% prediction accuracy in independent validation. This study not only clarified the quantitative dependence of the damage extent to typical SPUA microstructures on particle size reduction induced by fine grinding, but also proposed a novel SEM image analysis method based on machine learning. This approach provides an efficient and objective new pathway for quantitatively assessing the impact of the grinding process on the functional structures of SPUA modifiers.
扫码关注我们
求助内容:
应助结果提醒方式:
