An Evolutive-Deformation approach to enhance self-supporting areas in Additive Manufacturing designs

Abstract

Additive Manufacturing (AM) technology has matured rapidly in recent years, opening an engineering design freedom front when combined with Topology Optimization (TO). Although there is no doubt about their potential to provide optimal parts, there still exists a wide gap between them which reduces the efficiency of the design process. Overhang constraints in TO are still limited, therefore the designs generated by these algorithms are not yet ready for being directly manufactured. They usually need support structures to avoid droops or warps, which is time-consuming, costly and can lead to the loss of the optimal path since designers are required to manually redesign the geometry. The proposed approach combines both design and AM constraints into an automatic workflow in which the geometry provided by the TO software is evolved to minimize support structures for any printing direction while preserving its structural performance. The NSGA-II algorithm is coupled with a Free-Form Deformation (FFD) method to develop different geometries, while fitness is evaluated with FEM analysis. The framework provides Pareto Fronts with the most promising printing directions and geometries in terms of support structures, stiffness and weight. This multi-objective optimization approach that aims to improve TO parts manufacturability is applied to two case studies and one real-world part, highlighting on average a reduction of support structures of around 30%, whereas mass is practically preserved and stiffness decreased by 5%.