Topology optimisation for design and additive manufacturing of functionally graded lattice structures using derivative-aware machine learning algorithms

Abstract

Although additive manufacturing has offered substantially new opportunities and flexibility for fabricating 3D complex lattice structures, effective design of such sophisticated structures with desired multifunctional characteristics remains a demanding task. To tackle this challenge, we develop an inventive multiscale topology optimisation approach for additively manufactured lattices by leveraging a derivative-aware machine learning algorithm. Our objective is to optimise non-uniform unit cells for achieving an as uniform strain pattern as possible. The proposed approach exhibits great potential for biomedical applications, such as implantable devices mitigating strain and stress shielding. To validate the effectiveness of our framework, we present two illustrative examples through the dedicated digital image correlation (DIC) tests on the optimised samples fabricated using a powder bed fusion (PBF) technique. Furthermore, we demonstrate a practical application of our approach through developing bone tissue scaffolds composed of optimised non-uniform iso-truss lattices for two typical musculoskeletal reconstruction cases. These optimised lattice-based scaffolds present a more uniform strain field in complex anatomical and physiological condition, thereby creating a favourable biomechanical environment for maximising bone formation effectively. The proposed approach is anticipated to make a significant step forward in design for additively manufactured multiscale lattice structures with desirable mechanical characteristics for a broad range of applications.