Topology optimization of fiber reinforced plastic structures

Abstract

The utilization of carbon fiber reinforced polymer (CFRP) composites is rapidly expanding across a range of industries attributable to their exceptional material properties, including lightweight, high strength-to-weight ratio, stiffness-to-weight ratio, corrosion resistance, great designability, and excellent fatigue characteristics. Topology optimization and fiber path optimization of CFRP structures enable to further improve the ratios of strength/stiffness to weight for achieving a better lightweight design. In this study, design optimization of topological configuration and fiber paths are conducted using the gradient-based methods, in which design sensitivities are derived for all the optimization methods, including discrete material optimization (DMO) and level set method (LSM). To address the problem of the convergence to local optima in the gradient-based methods, a novel procedure combining the machine learning (ML) techniques with DMO is proposed to search for a quasi-global optimum. To accommodate the advanced manufacturing technologies such as additive manufacturing (also named 3D printing), LSM is proposed for both topology and fiber path optimizations, which enables to present the filament paths inherently without further reconstruction of the nozzle path. Combing with the fast-marching method of LSM, manufacturing drawbacks of gaps and overlaps induced from the non-parallel filament paths can be addressed. Optimized and empirical samples are fabricated with 3D printers to demonstrate the effectiveness of the proposed new techniques for further improving the structural performances.