Multi-material mechanical metamaterials offer great potential for lightweight and multifunctional engineering structures, yet conventional manufacturing limits their complexity. Although additive manufacturing (AM) enables unprecedented design freedom, optimizing multi-material lattices with distinct mechanical properties remains challenging, leaving much of the design space unexplored. Here, we develop complex multi-material lattice structures that combine polylactic acid (PLA) for stiffness with thermoplastic polyurethane (TPU) for flexibility. Additive manufacturing is coupled with systematic optimization of key design parameters, including unit-cell configuration (2×, 3×, and 4×), inner material thickness, and outer wall thickness, using a Taguchi Design of Experiments and validated finite element analysis (FEA). Quasi-static compression testing shows that increasing the unit-cell count significantly improves plateau-region stability for both lattice types. For the strut-based lattices, the 4× configuration achieves a 78% increase in normalized peak load and a 57% increase in energy absorption. In contrast, the plate-based lattices exhibit improvements of 42% in peak load and 80% in energy absorption at the same unit-cell configuration. FEA validated the experimental outcomes, revealing consistent force-deformation behavior. Scanning electron microscopy (SEM) analysis provided further insights, revealing excellent interfacial bonding between TPU and PLA, along with distinct fracture mechanisms between strut-based and plate-based structures. Additionally, crashworthiness analyses highlighted a consistent upward trend in key metrics such as mean crushing force (MCF) and crush force efficiency (CFE) with increased unit cell counts and outer wall thickness. CFE values lie between the desired range of 80 to 97% in nearly all designs, except the strut-based 2x cell arrangement. Ultimately, this study highlights additive manufacturing’s potential to engineer multi-material metamaterials tailored for aerospace, biomedical, and other high-performance applications.
Abbas et al. (2026) studied this question.