Phase topology and morphology of co-continuous tungsten‑copper composites

The functional and mechanical properties of co-continuous tungsten–copper (W Cu) composites are strongly influenced by the topology and morphology of their interpenetrating phases. However, quantitative microstructural characterization remains challenging due to limitations of conventional techniques. This study combines TriBeam tomography with deep learning-based image segmentation to reconstruct and analyze the three-dimensional microstructures of two W Cu composites with distinct phase compositions (58 W–42Cu and 81 W–19Cu by volume). Key morphological and topological descriptors—including particle size, coordination number, shape, and junction area fraction—are extracted to elucidate inter-parameter relationships and gain insight into composite behavior. The experimentally derived data serve as a benchmark for evaluating synthetic microstructures generated computationally to replicate the observed phase topology. Results show that tungsten coordination numbers remain consistent across compositions (~9), while junction area fractions differ by a factor of two. Strong pairwise correlations are observed among coordination number, particle size, and shape. Comparisons between real and synthetic microstructures highlight both the strengths and limitations of computational models in capturing these distributions and correlations. Overall, this study provides a quantitative foundation for designing representative volume elements (RVEs) that accurately reflect co-continuous composite topology, with important implications for property prediction. • 3D TriBeam tomography used to map W Cu composite phase topology. • Deep learning segmentation enables quantitative phase separation. • Coordination, shape, and junction metrics extracted from 3D data. • Synthetic microstructures benchmarked against experimental topology. • Framework established for realistic RVE design in co-continuous systems.