Forming process optimisation for variable geometries by machine learning – Convergence analysis and assessment

Forming process optimisation for variable geometries by machine learning – Convergence analysis and assessment

ZIMMERLING Clemens, KÄRGER Luise

download PDF

Abstract. For optimum operation, modern production systems require a careful adjustment of the employed manufacturing processes. Physics-based process simulations can effectively support this process optimisation; however, their considerable computation times are often a significant barrier. One option to reduce the computational load is surrogate-based optimisation (SBO). Although SBO generally helps improve convergence, it can turn out unwieldy when the optimisation task varies, e.g. due to frequent component adaptations for customisation. In order to solve such variable optimisation tasks, this work studies how recent advances in machine learning (ML) can enhance and extend current surrogate capabilities. More specifically, an ML-algorithm interacts with generic samples of component geometries in a forming simulation environment and learns to optimise a forming process for variable geometries. The considered example of this work is blank holder optimisation in textile forming. After training, the algorithm is able to give useful recommendations even for new, non-generic geometries. While the prior work considered initial recommendations only, this work studies the convergence behaviour upon component-specific algorithm refinement (optimisation) at the example of two geometries. The convergence of the new pre-trained ML-approach is compared to classical SBO and a genetic algorithm (GA). The results show that initial recommendations indeed converge to the process optimum and that the speed of convergence outperforms the GA and compares roughly to SBO. It is concluded that – once pretrained –the new ML-approach is more efficient on variable optimisation tasks than classical SBO.

Keywords
Forming Optimisation, Surrogate, Machine Learning, Variable Geometries

Published online 4/19/2023, 12 pages
Copyright © 2023 by the author(s)
Published under license by Materials Research Forum LLC., Millersville PA, USA

Citation: ZIMMERLING Clemens, KÄRGER Luise, Forming process optimisation for variable geometries by machine learning – Convergence analysis and assessment, Materials Research Proceedings, Vol. 28, pp 1155-1166, 2023

DOI: https://doi.org/10.21741/9781644902479-126

The article was published as article 126 of the book Material Forming

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 license. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

References
[1] N. Shamsaei, A. Yadollahi, L. Bian, S.M. Thompson, An overview of direct laser deposition for additive manufacturing – part 2, Additive Manuf. 8 (2015) 12-35. https://doi.org/10.1016/j.addma.2015.07.002
[2] L. Kärger, A. Bernath, F. Fritz, S. Galkin, D. Magagnato, A. Oeckerath, A. Schön, F. Henning, Development and validation of a cae chain for unidirectional fibre reinforced composite components, Compos. Struct. 132 (2015) 350-358. https://doi.org/10.1016/j.compstruct.2015.05.047
[3] S. Koziel, L. Leifsson, Surrogate-Based Modeling and Optimization: Applications in Engineering. Springer Science+Business Media, New York, 2013. https://doi.org/10.1007/978-1-4614-7551-4
[4] C. Zimmerling, C. Poppe, O. Stein, L. Kärger, Optimisation of manufacturing process parameters for variable component geometries using reinforcement learning, Mat. Des. 214 (2022) 110423. https://doi.org/10.1016/j.matdes.2022.110423
[5] H.R. Attar, H. Zhou, A. Foster, N. Li, Rapid feasibility assessment of components to be formed through hot stamping: A deep learning approach, J. Manuf. Process. 68 (2021) 1650-1671. https://doi.org/10.1016/j.jmapro.2021.06.011
[6] H. Zhou, Q. Xu, Z. Nie, N. Li, A study on using image-based machine learning methods to develop surrogate models of stamp forming simulations, J. Manuf. Sci. Eng. 144 (2022) 021012. https://doi.org/10.1115/1.4051604
[7] R.T. Haftka, D. Villanueva, A. Chaudhuri, Parallel surrogate-assisted global optimization with expensive functions – a survey, Struct. Multidisc. Optimization 54 (2016) 3-13. https://doi.org/10.1007/s00158-016-1432-3
[8] H. El Kadi, Modeling the mechanical behavior of fiber-reinforced polymeric composite materials using artificial neural networks – a review, Compos. Struct. 73 (2006) 1-23. https://doi.org/10.1016/j.compstruct.2005.01.020
[9] C. Zimmerling, C. Poppe, L. Kärger, Estimating optimum process parameters in textile draping of variable part geometries – A reinforcement learning approach, Procedia manuf. 47 (2020) 847-854. https://doi.org/10.1016/j.promfg.2020.04.263
[10] C. Zimmerling, D. Trippe, B. Fengler, L. Kärger, An approach for rapid prediction of textile draping results for variable composite component geometries using deep neural networks, AIP Conference Proceedings 2113, ESAFORM 2019 in Vitoria-Gasteiz/Spain (2019), Art. 020007. https://doi.org/10.1063/1.5112512
[11] S. Haanappel, Forming of UD fibre reinforced thermoplastics: A critical evaluation of intra-ply shear. PhD-Thesis, Universiteit Twente, 2013. https://doi.org/10.3990/1.9789036535014
[12] Sandia National Laboraties. DAKOTA user’s manual – version 6.15, www.dakota.sandia.gov/sites/default/files/docs/6.15/Users-6.15.0.pdf (2021), last accessed: 31.08.2022.
[13] C. Zimmerling, P. Schindler, J. Seuffert, L. Kärger: Deep neural networks as surrogate models for time-efficient manufacturing process optimisation. PoPuPS of ULiège Library, ESAFORM 2021 in Liège/Belgium (2021). https://doi.org/10.25518/esaform21.3882