Generative Design
Gu Research Group - UC Berkeley
Gu Research Group - UC Berkeley
Birn, B., Woody, K.*, Sun, D.*, Gu, G. X. (2025). Advanced Engineering Materials, 2025, 0, e202501559.
*indicates equal contribution
Lattice structures can enhance mechanical properties while minimizing mass, but often face challenges from inefficient material distribution and stress concentrations. Here, a generative design method is used to create freeform lattices that resemble biological structures. This approach is found to provide less constrained material redistribution, allowing the reduction of stress concentrations embedded in conventional designs. Three lattice types are optimized and compared: the bending-dominated body-centered cubic (BCC) lattice, the stretching-dominated simple cubic lattice, and a directional, water-lily-inspired lattice. Compression testing reveals improvements in stiffness, strength, and energy absorption in all three lattice types at sufficiently high relative density. Interestingly, all optimized structures display a marked reduction in anisotropy, with an optimized BCC lattice exhibiting isotropic elasticity. This study shows how generative design can emulate the organic forms of nature to create lattices with superior properties, offering new pathways for lightweight, high-performance structures.