ABSTRACT
Lattice materials are widely used in industries due to their designable capabilities of specific stiffness and energy absorption. However, evaluating the mechanical response of macroscopic lattice structures can be computationally expensive. Homogenization-based multi-scale analysis offers an efficient approach to address this issue. To achieve a simpler, while precise, homogenization, the authors proposed an equidistant segmentation (ES) method for the measurement of the effective shear modulus. In this method, the periodic boundary conditions (PBCs) are approximated by constraining the lateral displacement of nodes between parallel layers of periodic cells. The validations were applied to three typical lattice topologies: body-centered cubic (BCC) lattices, gyroid-, and primitive-triply periodic minimal surface (TPMS) lattices, to predict and compare their anti-vibration capacities. The results demonstrated the rationality and the promising precision of the multi-scale-based equivalent modal analysis through the proposed method and that it eliminated the geometric limitation of lattices with diverse frameworks. Overall, a higher anti-vibration capacity of TPMS was observed. In the study, the authors examined the influence of the relative densities on the balance between the anti-vibration capacity and loading capacity (per unit mass) of the TPMS topologies. Specifically, the unit mass of the TPMS with lower relative densities was able to resist higher frequencies, and the structures were dominated by the anti-vibration capacity. In contrast, a higher relative density is better when emphasizing the loading capacity. These findings may provide notable references to the designers and inform the selection of lattice materials for various industrial applications.
ABSTRACT
Reliable extreme lightweight is the pursuit in many high-end manufacturing areas. Aided by additive manufacturing (AM), lattice material has become a promising candidate for lightweight optimization. Configuration of lattice units at the material level and the distribution of lattice units at the structure level are the two main research directions recently. This paper proposes a generative strategy for lattice infilling optimization using organic strut-based lattices. A sphere packing algorithm driven by von Mises stress fields determines the lattice distribution density. Two typical configurations, Voronoi polygons and Delaunay triangles, are adopted to constitute the frames, respectively. Based on finite element analysis, a simplified truss model is utilized to evaluate the lattice distribution in terms of mechanical properties. Optimization parameters, including node number, mapping gradient, and the range of varying circle size, are investigated through the genetic algorithm (GA). Multiple feasible solutions are obtained for further solidification modelling. To avoid the stress concentration, the organic strut-based lattice units are created by the iso-surface modelling method. The effectiveness of the proposed generative approach is illustrated through a classical 3-point bending beam. The stiffness of the optimized structure, verified through experimental testing, has increased 80% over the one using the traditional uniform body center cubic (BCC) lattice distribution.
