3D shape quantification and random packing simulation of rock aggregates using photogrammetry-based reconstruction and discrete element method

Rock aggregates are commonly used granular material in building engineering. The 3D reconstruction of realistic rock aggregates is significant for the study of their mechanical properties in numerical simulation. This paper aims to create an efficient particle reconstruction system and analyse the effects of particle shapes on the random packing properties of rock aggregates. First, a photographic system is established to obtain 2D images of rock aggregates from multiple shooting directions, and the photogrammetric technology is employed to reconstruct the 3D model of the rock aggregate. A total of 1000 realistic 3D models of rock aggregates are obtained. Second, four shape indexes, i.e., elongation index (EI), flatness index (FI), convexity index (CI), and roundness (RD), are employed to characterize the shapes of the obtained rock aggregates. The results of statistical analysis show that the probability distribution of rock aggregates’ shape indexes can be well fitted by Gaussian function. Third, the initial samples with different EI and FI values are generated through a series of computational algorithms based on MATLAB and then are imported into PFC^3D for discrete element simulation. Subsequently, random packing of rock aggregates is carried out by two different methods, i.e., the gravity falling method (GFM) and the isotropic compression method (ICM). The results show that the EI and FI have significant influences on packing density, coordination number, fabric anisotropy, and contact force between the samples.