A novel surrogate modelling approach for additive manufacturing processes

Abstract

AbstractAdditive manufacturing (AM) enables the production of complex and customised components with minimal waste. However, to achieve the desired geometry and quality, a lot of process optimisation is needed (often by experimental trial and error). Especially, in the case of AM with concrete materials, the time‐dependent behaviour of the fresh concrete is a large challenge to achieve process stability. Simulations taking into account the time‐dependent material are able to predict the printing result. Since these simulations are very time consuming, a proper orthogonal decomposition (POD)‐based reduced order model for AM is presented. Based on snapshots along the build‐up process, a surrogate model is constructed that is able to predict the printing result with respect to the underlying process and materials. Inspired by the POD approximation, a novel surrogate modelling approach using the initial waypoints of the printing trajectory as input is investigated. In this case, it is possible to analyse the manufacturing process for arbitrary geometries. The proposed method is demonstrated for the use case of additive manufacturing in construction.