A Two-Dimensional Sequential Packing Method for Lunar Regolith Particles Based on Random Polygons

To accurately characterize the effects of polydisperse particle sizes, multimineral composition, and angular morphology on the packing structure of lunar regolith, a two-dimensional sequential packing method based on random convex octagons is proposed. The method establishes a particle parameter system using data from Chang’e-5 samples and generates polygonal particle models with controllable angular features through radial perturbation. On this basis, a sequential packing algorithm based on available arc analysis is developed. Non-overlapping particle insertion is achieved via geometric envelope constraints, and progressive filling is realized through effective arc sampling. Meanwhile, a packing control coefficient is introduced to enable continuous regulation of packing density. Results show that the proposed method can generate highly dense particle assemblies, with a maximum packing density of 0.8757 and an average coordination number of approximately 3.18, capturing the structural characteristics of “high compactness–low coordination number” in polydisperse angular particle systems. The algorithm exhibits a computational complexity of O(N^1.628), demonstrating high efficiency. Furthermore, contact area and contact strength are quantitatively characterized through contact contour extraction and an equivalent bow-shaped model. Radial distribution function and contact statistics indicate that the generated structures possess good randomness and physical consistency. The proposed method provides a high-fidelity mesoscopic structure generation approach for discrete element modeling (DEM) of lunar regolith and establishes a reliable foundation for analyzing the mechanical behavior of granular systems.