This work presents an efficient algorithm for generating statistically representative microstructures of particulate composites in periodic representative volume elements. The Swelling and Random Migration (SRM) algorithm combines collective particle rearrangements with an adaptive cell-based neighbor-search scheme, enabling near-linear computational scaling for low to intermediate volume fractions and allowing simulations with up to $10^7$ particles in two and three dimensions. SRM offers great flexibility, allowing the controlled generation of both equilibrium-like and strongly non-equilibrium particle arrangements. The method is readily extendable to non-spherical inclusions; this capability is demonstrated by modeling thin circular platelets and generating qualitatively distinct platelet microstructures, including highly interconnected "house-of-cards" networks and metastable quasi-nematic domains. The results highlight the importance of microstructural arrangement in structure-property relationships and establish SRM as a powerful tool for generating realistic, diverse, and computationally accessible particle configurations for composite material modeling.

翻译：本文提出了一种高效算法，用于在周期性代表体元中生成具有统计代表性的颗粒复合材料微结构。溶胀与随机迁移（SRM）算法将集体粒子重排与自适应胞元邻域搜索方案相结合，在低至中等体积分数下实现近线性计算扩展，并支持在二维和三维空间中模拟多达$10^7$个粒子。SRM算法具有极高的灵活性，可受控生成类平衡态与强非平衡态两种粒子排列模式。该方法可便捷扩展至非球形夹杂物：通过模拟薄圆片状血小板，我们展示了该能力，并生成了性质截然不同的血小板微结构，包括高度互联的"纸牌屋"网络与亚稳态准向列域。研究结果揭示了微结构排列在结构-性能关系中的关键作用，并确立了SRM算法作为复合材料建模中生成真实、多样且计算可行的粒子构型的强大工具。