In-mould assembly of functionally integrated structures: A surrogate model for fast quality assessment

Abstract

For the in-mould assembly of functionally integrated structures, a proper bond strength development between the different components has to be ensured during the actual manufacturing step. Physically detailed models and simulations are able to compute a reliable bond strength distribution. However, these computations are very time consuming. In this contribution, a simulation based surrogate model for the in-mould assembly of thermoplastic composite structures is developed. It predicts the bond strength within seconds using data from process simulations combined with machine learning, achieving a remaining error of less than 5% compared to high-fidelity computations.