Thermal and Mechanical Homogenization of Architected Interpenetrating Composites using a Voxel-Based Fast Fourier Solver
Please login to view abstract download link
Recent advancements in additive manufacturing processes allow for the fabrication of complex interpenetrating metal-metal hybrid composite structures that are promising candidates to satisfy the growing demand for new and improved engineering materials [1]. Previous research was primarily focused on single-phase structures or polymer-polymer and polymer-ceramic composites, respectively. The combination of different metals within a single part opens up a multitude of new potential application areas, ranging from heat exchangers to welding connectors or parts with improved energy absorption. However, critical regions, such as localized stress concentrations, can occur under load, especially near the interface between the included phases. In order to design complex hybrid composite structures with tailored material properties, a deep understanding of the underlying thermal and mechanical behavior is necessary. We present a computationally efficient workflow for generation and numerical analysis of the structure-property relationship of interpenetrating metal-metal composites based on triply-periodic minimal surface lattice structures. We propose systematically generated voxel-based microstructure models with different topologies and high flexibility in volume fractions. A homogenization approach based on a state-of-the-art Fast Fourier Transform (FFT) solver is used to evaluate homogenized thermal and mechanical properties such as effective thermal conductivity or effective stiffness. By analyzing, we compare different variations of composite structures and identify promising candidates for selected load scenarios. Based on these results, the effective structure-property relations of the hybrid metal-metal composites are discussed. This workflow lays the foundation for a parameter-based optimization and could be adapted to other material systems in the future. Due to the high computational efficiency of FFT-based methods, it is suitable to build a large database that can be used for machine learning and AI tasks. Furthermore, the workflow can be extended to study the behavior of more complex microstructures, including functional gradients. [1] Zuyao Zhang, Zhi Wang, Qizhong Zhao, Konda Gokuldoss Prashanth, Metal-metal interpenetrating phase composites: A review, Journal of Alloys and Compounds, 1009, 176951, 2024.
