Multiscale lattice structures lend themselves to optimisation problems as their global material properties can be tailored by modifying the configuration and arrangement of the underlying periodic microstructures. Due to their complex designs, lattice structures are typically manufactured using novel additive manufacturing (AM) techniques. However, the intricate designs of additively manufactured multiscale lattice parts often contain defects, leading to a reduction in their mechanical performance. This is particularly detrimental for optimised multiscale designs as their performance can be highly sensitive to any small changes. Therefore, to enable a safe and reliable application of multiscale lattice structures in industrial applications, it is important to consider the non-negligible impacts of irregularities that occur during the AM process.
To account for manufacturing irregularities a design dependent random defect model is introduced within the mutliscale optimisation framework. The model perturbs the radii of the microstructures, simulating weaker lattice structures caused by the random irregularities that occur during AM processes. Utilising this model within a robust optimisation framework provides the optimiser with the information required to design against the manufacturing uncertainties, generating structures that are significantly less sensitive to random design irregularities introduced during their manufacture.