Multiscale Optimization of Weav3D Lattice-Reinforced Composites for Lightweight Structural Design

Abstract

This thesis evaluates the effectiveness of conventional optimization methods applied to lattice-reinforced composite structures, specifically panels manufactured using Weav3D technology. Weav3D integrates thermoplastic composite tapes into tailored lattice geometries, enabling strategic reinforcement placement. Material properties of the carbon fiber tapes were first experimentally characterized through mechanical testing on a universal testing machine (UTM), with strain fields measured using Digital Image Correlation (DIC). These validated properties served as inputs to a computational optimization framework using Altair HyperStudy for parameter exploration, Jpanel for generating homogenized orthotropic material models, and OptiStruct for finite element analysis. A Design of Experiments (DOE) based on Latin Hypercube Sampling (LHS) guided surrogate modeling and global optimization. Results showed a consistent preference for warp-direction reinforcement under bending loads, with fill-direction material significantly reduced. The final optimized lattice-based configuration achieved a 21% reduction in mass compared to a continuous baseline panel, confirming that classical surrogate-based optimization techniques can effectively streamline composite design without compromising performance.