A Density-Based Topology Optimisation Method Including Geometrical Uncertainties
Please login to view abstract download link
In the last decade, new topology optimisation (TO) algorithms have been proposed, with special features associated with Additive Manufacturing (AM) processes. However, AM is affected by a lack of repeatability: the integration of uncertainty is thus important. In this article, the integration of uncertainties that affect the part geometry in TO algorithms is considered. On the one hand, the proposed approach is based on a classic density-based TO algorithm. On the other hand, the uncertainty of geometrical features is handled through process-related variables. Furthermore, uncertainty characterisation of AM processes shows the need to integrate epistemic uncertainties as aleatory uncertainty. In this context, the mathematical framework related to uncertainty is based on Dempster-Shafer or probability boxes structures, instead of the classical probability one. The effectiveness of the approach is tested on various benchmark structures taken from the literature. Obtained geometries are consistent with the robust design approach, thanks to the integration of AM process-related variables into the topology optimisation process.
