Data-Driven Optimization of UV-Curable Resin-Based Polymeric Lattice Structures for Predicting Load-Bearing Capacity and Structural Dimensions

Lattice structures fabricated by additive manufacturing offer significant potential for lightweight, high-performance load-bearing applications; however, identifying optimal lattice geometries for target and maximum mechanical performance remains challenging due to the high cost and time associated with extensive experimental testing. In this study, a novel data-driven optimization framework is proposed to design lattice structures with targeted compressive load-bearing capacity efficiently. Using additive manufacturing, lattice structures with controlled geometric parameters and predefined topologies were fabricated through a layer-by-layer 3D printing process. This approach enables precise control over lattice architecture and repeatable fabrication of complex cellular geometries. A total of 93 lattice specimens representing five different lattice topologies were additively manufactured and experimentally evaluated under compression testing. The resulting dataset, comprising lattice topology, geometric dimensions, and ultimate compressive load, was used to develop surrogate models based on Gradient Boosted Regression. Differential Evolution optimization algorithm to identify optimal lattice dimensions that can achieve specified target and maximum loads within predefined design constraints. The results demonstrate that the proposed hybrid framework effectively captures the nonlinear relationship between lattice geometry and mechanical performance, achieving both target and maximum load-bearing capacity, while enabling accurate and efficient optimization of lattice dimensions and significantly reducing experimental effort.