Proceedings of the XMO Industrial Seminar 2026: Excellence in Manufacturing and Operations
Keywords
constant-mean-curvature surfaces; triply periodic minimal surfaces; topology hybridisation; genus grading; nacre-inspired laminates; physics-informed Bayesian neural network; Mechanics of Structure. Genome homogenisation; multi-objective Bayesian optimisation; desktop additive manufacturing; multifunctional mechanical metamaterials
Tracks
MULTIFUNCTIONAL AND RESILIENT DESIGNS FOR MANUFACTURING
DOI
10.5703/1288284318687
Abstract
Architected solids that combine high stiffness, strength, toughness, and ductility in one geometry are difficult to manufacture without specialised feedstocks or sequential consolidation. Two complementary architectures are built here from a single curvature-programmed building-block library of four constant-mean-curvature (CMC) shell topologies — Schwarz Primitive (SP), Schwarz Diamond (SD), Schoen's I-WP, and FRD — that share commensurate cubic facets and admit C0 /C1 /C2 continuity at every interface. Multi-objective hybridisation gives an open-cell metamaterial with specific axial stiffness E*/ρ* = 63.6 GPa·cm3 ·g-1 and specific peak compressive stress σpeak/ρ* = 416.9 MPa·cm3 ·g-1, exceeding the I-WP single family by 40% in specific stiffness and 48% in specific strength, and four structural alloys by ~2.5× in specific stiffness and 2– 16× in specific strength at matched mass. Genus-grading the same library through SiC-loaded PLA bricks bonded by a hybrid-CMC PC/CF mortar interchanges the laminate's mechanical profile by changing only the slicing layer order. Across six architecture tiers (Solid → TPMS → CMC → Hybrid CMC → Gradient CMC → Nacre stack) strength, stiffness, fracture toughness, and ductility climb 9.1×, 9.7×, 8.6×, and 7.0× without any property trade-off. Both architectures are designed by one physics-informed Bayesian neural network coupled to Mechanics-ofStructure-Genome homogenisation and produced on commercial desktop additive-manufacturing hardware, opening a low-cost, scalable route to multifunctional parts.
Programmable Hybrid and Gradient Constant Mean Curvatures for High-Performance Multifunctional Mechanical Metamaterials
Architected solids that combine high stiffness, strength, toughness, and ductility in one geometry are difficult to manufacture without specialised feedstocks or sequential consolidation. Two complementary architectures are built here from a single curvature-programmed building-block library of four constant-mean-curvature (CMC) shell topologies — Schwarz Primitive (SP), Schwarz Diamond (SD), Schoen's I-WP, and FRD — that share commensurate cubic facets and admit C0 /C1 /C2 continuity at every interface. Multi-objective hybridisation gives an open-cell metamaterial with specific axial stiffness E*/ρ* = 63.6 GPa·cm3 ·g-1 and specific peak compressive stress σpeak/ρ* = 416.9 MPa·cm3 ·g-1, exceeding the I-WP single family by 40% in specific stiffness and 48% in specific strength, and four structural alloys by ~2.5× in specific stiffness and 2– 16× in specific strength at matched mass. Genus-grading the same library through SiC-loaded PLA bricks bonded by a hybrid-CMC PC/CF mortar interchanges the laminate's mechanical profile by changing only the slicing layer order. Across six architecture tiers (Solid → TPMS → CMC → Hybrid CMC → Gradient CMC → Nacre stack) strength, stiffness, fracture toughness, and ductility climb 9.1×, 9.7×, 8.6×, and 7.0× without any property trade-off. Both architectures are designed by one physics-informed Bayesian neural network coupled to Mechanics-ofStructure-Genome homogenisation and produced on commercial desktop additive-manufacturing hardware, opening a low-cost, scalable route to multifunctional parts.