Topology Optimization and Generative Design in Additive Manufacturing: A Systematic Review of Algorithms, Tools and Industrial Applications

Abstract

Additive manufacturing (AM) has increased the design space of engineering components, allowing the creation of geometries which are hard or even impossible to realise through conventional machining, casting and forming. The capacity has added to the curiosity surrounding generative design (GD) and topology optimisation (TO), which aim to enhance structural performance, minimise material consumption and investigate other design options. This review combines the technical compilation provided with the relevant scholarly literature on the relationship between TO, GD and AM. The review is organised around the reporting principles of PRISMA 2020 and focuses specifically on the review question, eligibility logic, organisation of the evidence, methodological limitations and research gaps. Literature suggests that density-based approaches like Solid Isotropic Material with Penalisation (SIMP), level-set methods and evolutionary techniques continue to be important TO approaches, and GD is a broader requirements-driven approach which can involve optimisation, simulation, materials selection, and manufacturing constraints. The key result here is that having the best solution in the world of computation does not guarantee the best solution in the world of manufacturing, economy and reliability. All of these, such as overhangs, support structure, build orientation, enclosed voids, minimum feature size, anisotropy, residual stress, thermal distortion, surface finish, post-processing and cost, need to be taken into consideration in the design process. Applications in the aerospace, automotive and lightweight cellular structures sectors show promising potential, but benefits have only been reported in the context of specific applications. The next step in research should focus on coupled design–process optimisation, multi-physics modelling, uncertainty quantification, explainable artificial intelligence, experimental validation and closed-loop digital manufacturing. The synthesis is also embedded in the setting of an industrial economy, where knowledge and skills about the processes, their parameters and constraints determine what can be practically achieved, and in which limited use of AM and simulation tools is prevalent. The review finds that the maturity of TO and GD should be assessed based on the quality of validated engineering outcomes and not just on the geometric complexity or weight reduction alone.