Paper Titles
On-Line Prediction of Dry Zones during Composite Process through Digital Shadow Approach
p.123
Graph Neural Network for Draw-in Prediction in Sheet Metal Stamping
p.137
A Bayesian Data Assimilation Framework for Characterizing Recrystallization in Steels
p.145
Predicting Lateral Flow in Hot Sheet Metal Rolling Using Symbolic Regression
p.157
Physics-Informed Recurrent Neural Networks with Kinematic Constraints for Large Deformation Metal Forming
p.171
An Ai-Based Approach to Developing a Microstructural Model for Multi-Stage Hot Deformation Processes
p.183
Accelerating Microstructural Evolution Simulations in DIGIMU® through a Front-Tracking Lagrangian Solver: Implementation and Validation in AISI 304L Stainless Steel
p.193
Study on the Apparent Friction Coefficient between a Deformable and a Rough Rigid Body under Large Sliding
p.205
Using Kolmogorov-Arnold Networks for Discrepancy Modeling of Lateral Flow in Hot Rolling of Steel Slabs
p.215

Physics-Informed Recurrent Neural Networks with Kinematic Constraints for Large Deformation Metal Forming

Article Preview
View Pdf By email

Abstract:

Metal-forming is a manufacturing process that involves non-linear elastoplastic deformations to shape a blank into a complex geometry. These processes are governed by numerous parameters, with significant influence on the final product. To analyse the effects of such parameters, large-scale finite element (FE) simulations are often conducted. However, these models are computationally expensive and often unsuitable for real-time analysis. To overcome these limitations, surrogate models have emerged as powerful alternatives. In this study, we propose a physics-informed recurrent neural network framework (PIRNN) to evaluate displacements and strain tensor components. In particular, the latter are not a network output but are obtained through the application of kinematic relations. Given the initial configuration as input and the final configuration as output, it is possible to evaluate the deformation gradient. The impenetrability condition is then injected into the loss to improve the estimation of displacements and strain tensors. The PIRNN model, referred to as a kinematics-informed recurrent neural network, is trained on data generated from FE simulations of a deep-drawing process. The accuracy of the model is evaluated on a test dataset (design points that the model does not see during training) using different error measures. The results show that the proposed KI-RNN model can fairly reproduce the FE simulation results fairly well.

You have full access to the following eBook
Advanced Numerical and AI Strategies for Material Forming Read eBook

Info:

Periodical:

Key Engineering Materials (Volume 1050)

Pages:

171-182

DOI:

https://doi.org/10.4028/p-ls5PTZ

Citation:

Cite this paper

Online since:

April 2026

Authors:

Francesco Munzone*, Javad Hazrati, Wouter Hakvoort, Redmer van Tijum, Ton van den Boogard

Keywords:

Deep-Drawing, Design Exploration, Physics-Informed, Recurrent Neural Network

Export:

RIS, BibTeX

Permissions:

Creative Commons CC BY 4.0

Share:

Citation:

* - Corresponding Author

References

[1] S. Niu, E. Zhang, Y. Bazilevs and V. Srivastava, "Modeling finite-strain plasticity using physics-informed neural network and assessment of the network performance", J. Mech. Phys. Solids, vol. 172, no. 0022-5096, p.105177, 2023.

DOI: 10.1016/j.jmps.2022.105177

Google Scholar

[2] M. Maia, I. Rocha, D. Kovačević and F. v. d. Meer, "Physically recurrent neural networks for path-dependent heterogeneous materials: Embedding constitutive models in a data-driven surrogate", Comput. Methods Appl. Mech. Eng., vol. 407, no. 0045-7825, p.115934, 2023.

DOI: 10.1016/j.cma.2023.115934

Google Scholar

[3] M. Maia, I. Rocha, D. Kovačević and F. v. d. Meer, "Physically recurrent neural network for rate and path-dependent heterogeneous materials in a finite strain framework", Mech. Mater., vol. 198, no. 0167-6636, p.105145, 2024.

DOI: 10.1016/j.mechmat.2024.105145

Google Scholar

[4] J. C. Simo and T. J. R. Hughes, Computational Inelasticity, New York: Springer, 1998.

Google Scholar

[5] Li, W. W. and C. F. Jeff Wu, "Columnwise-Pairwise Algorithms with Applications to the Construction of Supersaturated Designs.," Technometrics, vol. 2, p.171–179, 1997.

DOI: 10.1080/00401706.1997.10485082

Google Scholar

[6] K. Q. Ye, W. Li and A. Sudjianto, "Algorithmic construction of optimal symmetric Latin hypercube designs", J. Stat. Plan. Inference , vol. 90, no. 0378-3758, pp.145-159, 2000.

DOI: 10.1016/s0378-3758(00)00105-1

Google Scholar

[7] R. Jin, W. Chen and A. Sudjianto, "An efficient algorithm for constructing optimal design of computer experiments", J. Stat. Plan. Inference, vol. 134, no. 1, pp.268-287, 2005.

DOI: 10.1016/j.jspi.2004.02.014

Google Scholar

[8] Y. G. Saab and V. B. Rao, "Combinatorial optimization by stochastic evolution", IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst., vol. 10, no. 4, pp.525-535, 1991.

DOI: 10.1109/43.75636

Google Scholar

[9] M. D. Morris and T. J. Mitchell, "Exploratory designs for computational experiments", J. Stat. Plan. Inference, vol. 43, no. 3, pp.381-402, 1995.

Google Scholar

[10] M. Veldhuis, J. Heingärtner, A. Krairi, D. Waanders and J. Hazrati, "An Industrial-Scale Cold Forming Process Highly Sensitive to Temperature Induced Frictional Start-up Effects to Validate a Physical Based Friction Model", Procedia Manuf., vol. 47, no. 2351-9789, pp.578-585, 2020.

DOI: 10.1016/j.promfg.2020.04.176

Google Scholar

[11] R. Jin, W. Chen and S. T. W., "Comparative studies of metamodelling techniques under multiple modelling criteria", Struct. Multidiscip. Optim., vol. 23, no. 1, pp.1-13, 2001.

Google Scholar

[12] L. N. Smith, "Cyclical Learning Rates for Training Neural Networks", in IEEE WACV, 2017.

Google Scholar

[13] K. Cho, B. van Merrienboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk and Y. Bengio, "Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling", arXiv, 2014.

DOI: 10.3115/v1/d14-1179

Google Scholar

[14] K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition", in IEEE CVPR, 2016.

Google Scholar