Delaunay‐Guided Random Sequential Expansion for Efficient Microstructure Generation in Fiber‐Reinforced Composites

ABSTRACT The random sequential expansion (RSE) method, known for its simplicity and ease of implementation, remains a widely adopted approach for generating random microstructures in unidirectional fiber‐reinforced composites. However, RSE is inherently limited by its inability to target and fill free spaces within the representative volume element (RVE), leading to suboptimal performance. In this study, Delaunay triangulation is used to detect free spaces within the RVE and guide fiber expansion fronts in the RSE method. The combination of Delaunay triangulation and RSE results in a simple, fast, and powerful algorithm that retains the advantages of RSE while overcoming its shortcomings. The effectiveness of the proposed method is demonstrated by generating RVEs for different fiber volume fractions and comparing its performance with previous methods, showing significantly shorter runtimes and a much higher success rate. A key advantage of the proposed approach for detecting filled and empty regions within the RVE is its remarkable flexibility. This adaptability allows for seamless integration not only with the RSE method but also with a wide range of existing fiber generation techniques, offering the potential to greatly enhance their performance and efficiency.