粗集料三维形态特征评价及基于其形态指标的沥青混合料性能优化研究

The significant effect of coarse aggregate morphology on asphalt mixture performances has been demonstrated by studies at home and abroad. Two-dimensional (2D) image is now usually used by researchers to study the coarse aggregate morphology. And the coarse aggregate morphological indices have not been introduced into the design of asphalt mixture to improve its performances. In view of the defect of 2D image method and the research need, this project proposes to study the methods for characterizing the three-dimensional (3D) morphology of coarse aggregates and study the method for optimizing the performances of asphalt mixture based on coarse aggregate morphological indices. The image acquisition system will be developed and the 3D reconstruction algorithms will be proposed to characterize the 3D shape and angularity of coarse aggregates. And the roughness tester will be introduced to characterize the 3D texture of coarse aggregates. The coarse aggregate morphological indices will then be introduced into the asphalt mixture design by means of gradation design. To optimize the performances of asphalt mixture, the objective function of gradation design and the equation of gradation design will be established to screen the coarse aggregate gradation and solve the optimum coarse aggregate gradation. The research results of this project will contribute to build the 3D index system of coarse aggregate morphology, which will lay the scientific basis for the quality inspection and quality control of coarse aggregates and for the performance improvement of asphalt mixture. Furthermore, the proposed method for optimizing the performances of asphalt mixture based on coarse aggregate morphological indices will provide an effective technical approach to enhance the performances of asphalt mixture.

粗集料形态对沥青混合料性能的重要影响已被国内外的研究所证明。目前绝大部分研究使用二维图像评价粗集料的形态特征，且粗集料形态指标尚未被引入沥青混合料的设计中改进其性能。针对二维图像方法的缺陷和研究需求，本项目研究粗集料三维形态特征的评价方法和基于其形态指标的沥青混合料性能优化方法。通过研制图像采集系统和构建三维重构算法完成粗集料多视角二维图像的采集和三维图像的重建，实现形状和棱角的三维评价；通过引进粗糙度测试仪实现纹理的三维评价。以级配设计为桥梁，将粗集料形态指标引入沥青混合料的设计中，通过建立级配设计目标函数和级配设计方程筛选粗集料级配和求解最优级配，实现沥青混合料性能的优化。项目的研究成果有助于构建粗集料形态三维评价指标体系，为粗集料的质量检验、质量控制以及沥青混合料的性能优化奠定科学依据；所提出的基于粗集料形态指标的沥青混合料性能优化方法可为改善沥青混合料的性能提供有效的技术途径。

粗集料形态对沥青混合料性能的重要影响已经被国内外的研究所证明。在过去二十年中，对粗集料形态特征已经有了大量研究，但绝大部分都是基于粗集料二维图像的方法。本项目从粗集料形状、棱角和纹理的定义出发、立足现有技术和结合工程实际，提出研究粗集料形状、棱角和纹理的三维评价指标。通过高精度三维扫描仪实现了粗集料三维点云数据的采集，建立了粗集料三维样本库。通过Khachiyan算法，建立问题模型，得到满足包围粗集料所有点数据的椭球方程，然后通过线性规划的思想，利用凸优化方法求得最优解，从而找到体积最小的能够包围所有顶点的椭球。通过调整椭球重心，改变椭球轴长的方法得到拟合误差最小的椭球。基于粗集料点云数据，构建了粗集料棱角评价指标，并验证了该指标的合理性，解决了粗集料三维棱角表征问题。通过构建理想纹理几何体，对比研究了Laplacian平滑、距离加权Laplacian平滑、曲率加权Laplacian平滑及Taubin平滑算法在纹理平滑方面的效果，并基于此提出了改进的拉普拉斯平滑算法。针对改进的拉普拉斯平滑算法，系统研究了遍历范围、平滑界限、平滑因子以及网格精度对平滑结果的影响，提出了最优化的平滑参数组合，解决了粗集料纹理的三维表征问题。本项目的研究成果有助于构建粗集料形态三维评价指标体系，为粗集料的质量检验、质量控制以及沥青混合料的性能优化奠定科学依据。