Investigation on Three-Dimensional Void Mesostructures and Geometries in Porous Asphalt Mixture Based on Computed Tomography (CT) Images and Avizo

Abstract

To investigate the void mesostructure in porous asphalt mixtures (PA), computed tomography (CT) and Avizo were utilized to scan and reconstruct the three-dimensional (3D) void model of PA-16 specimens. The void mesostructure of the specimen was quantitatively characterized through the anisotropy evaluation index. The equivalent pore network model (PNM) was extracted using the medial axis method. Based on the PNM model, the topological structure of the specimen and the morphological characteristics of the connected pores were analyzed. The results showed that the void anisotropy evaluation method can reflect the microscopic morphology of voids in porous asphalt mixtures. The cross-sectional porosity of representative elementary volume (REV) is mainly distributed between 20% and 25%, and about 90% of the macropores have a diameter between 0.5 mm and 3 mm. The distribution of cross-sectional porosity is uneven along the REV height direction. As the smallest cross-section of the seepage path, the equivalent radius of the throat is mainly between 0.1 mm and 1.5 mm, which is much smaller than the equivalent radius of the pore. The topological structure of pores is quite different, and their coordination numbers are mainly concentrated within 18. The pores with coordination numbers 1 to 10 constitute the main body of the pores inside REV, accounting for over 98% of the total number of pores. In addition, the permeability calculation results show that there is a significant difference in the permeability of each axis of REV compared to the total permeability of the superpave gyratory compactor (SGC) specimen, which illustrates that the permeability distribution presents an obvious spatial anisotropy. This study effectively reveals the heterogeneity of the 3D void morphology of porous asphalt mixtures, and it provides a reference for a better understanding of the void flow rules in drainage pavements.

Multiphase Reconstruction of Heterogeneous Materials Using Machine Learning and Quality of Connection Function

“…The edges within this graph are established between each voxel's corresponding vertex, such as x, and all its neighbors, encompassing the vertices of neighboring voxels that share at least one edge with voxel x. To be more precise, this relationship is defined in Equation (8).…”

“…Over the past year, various initiatives have been undertaken to attain the threedimensional structure of materials [5][6][7][8]. The initial efforts focused on achieving the materials' structure through practical methodologies including microtomography and stochastic reconstruction of the microstructure [9][10][11].…”