Geometric optimization of two-stage stamping dies for ultra-thin titanium bipolar plates using Sequential Physics-Informed Neural Networks

Zijie Ke¹ , Yiwen Huang¹ , Ziqiang Guo¹ , Yao Xiao¹ *, Zeran Hou¹ , Junying Min²

¹School of Mechanical Engineering, Tongji University, Shanghai 201804, China.

²College of Automotive and Energy Engineering, Tongji University, Shanghai 201804, China.

*corresponding author

DOI:

https://doi.org/10.7494/cmms.2026.1.1038

Abstract:

Bipolar plates are critical core components in proton exchange membrane fuel cells (PEMFCs). Titanium-based materials are highly favored due to their excellent corrosion resistance and high specific strength. However, the plates often experience severe local thinning and poor consistency in forming dimensions during the two-stage stamping process. Although traditional finite element method (FEM) optimization can mitigate these defects, it comes with high computational costs and time consumption. This study proposes a die design optimization framework based on the Sequential Physics-Informed Neural Network (S-PINN). Unlike traditional single-layer neural network models, S-PINN adopts a sequential architecture that effectively maps the two-stage forming process of the plates. This architecture can explicitly predict the evolution of forming quality from the pre-forming stage to the final stage. By embedding the core physical laws of plastic deformation into the network loss function, the S-PINN model effectively predicts the complex nonlinear relationship between mold geometry and forming quality, while ensuring physical consistency. Experimental and simulation results show that the S-PINN model’s prediction accuracy for dimensional consistency (DC) is 73.8% higher than that of the PINN model and 33.9% higher than that of the S-ANN model. Compared with traditional modeling methods, the S-PINN-optimized die design can reduce the thinning rate and improve channel dimensional consistency.

Cite as:

Ke, Z., Huang, Y., Guo, Z., Xiao, Y., Hou, Z., & Min, J. (2026). Geometric optimization of two-stage stamping dies for ultra-thin titanium bipolar plates using Sequential Physics-Informed Neural Networks. Computer Methods in Materials Science, 26(1), – . https://doi.org/10.7494/cmms.2026.1.1038

Article (PDF):

Accepted Manuscript – final pdf version coming soon

Keywords:

Sequential Physics-Informed Neural Network (S-PINN), Ultra-thin titanium sheet, Two-stage forming, Die shape optimization, Thinning rate, Dimensional consistency

Publication dates:

Received: 06.03.2026, accepted: 13.04.2026, published: XX.04.2026

Publication type:

Original scientific paper

References:

Alhalaybeh, T. S., Muhamaiti, Y., Shang, H., Zhou, L., Liang, X., & Lou, Y. (2026). Artificial neural network based stamping process design for three-point bending. Acta Mechanica Sinica, 42(3), 424895. https://doi.org/10.1007/s10409-025-24895-x

Asuero, A. G., Sayago, A., & González, A. G. (2006). The correlation coefficient: An overview. Critical Reviews in Analytical Chemistry, 36(1), 41–59. https://doi.org/10.1080/10408340500526766

Bajda, S., & Krzyzanowski, M. (2019). Modelling aspects of laser cladding of bioactive glass coatings on ultrafine-grained titanium substrates. Computer Methods in Materials Science, 19(3), 138–149. https://doi.org/10.7494/cmms.2019.3.0637

Benesty, J., Chen, J., Huang, Y., & Cohen, I. (2009). Pearson correlation coefficient. In Noise Reduction in Speech Processing (pp. 1–4). Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-00296-0_5

Bezerra, M. A., Santelli, R. E., Oliveira, E. P., Villar, L. S., & Escaleira, L. A. (2008). Response surface methodology (RSM) as a tool for optimization in analytical chemistry. Talanta, 76(5), 965–977. https://doi.org/10.1016/j.talanta.2008.05.019

Bong, H. J., Lee, J., Kim, J. H., Barlat, F., & Lee, M.-G. (2017). Two-stage forming approach for manufacturing ferritic stainless steel bipolar plates in PEM fuel cell: Experiments and numerical simulations. International Journal of Hydrogen Energy, 42(10), 6965–6977. https://doi.org/10.1016/j.ijhydene.2016.12.094

Daud, W. R. W., Rosli, R. E., Majlan, E. H., Hamid, S. A. A., Mohamed, R., & Husaini, T. (2017). PEM fuel cell system control: A review. Renewable Energy, 113, 620-638. https://doi.org/10.1016/j.renene.2017.06.027

Fan, E. (2000). Extended tanh-function method and its applications to nonlinear equations. Physics Letters A, 277(4–5), 212–218. https://doi.org/10.1016/s0375-9601(00)00725-8

Guo, N., Zhang, X., Hou, Z., Wang, W., Yang, D., Min, J., Ming, P., & Zhang, C. (2023). Hot stamping of ultra-thin stainless steel sheets for bipolar plates. Journal of Materials Processing Technology, 317, 117987. https://doi.org/10.1016/j.jmatprotec.2023.117987

Haghighat, E., Raissi, M., Moure, A., Gomez, H., & Juanes, R. (2021). A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics. Computer Methods in Applied Mechanics and Engineering, 379, 113741. https://doi.org/10.1016/j.cma.2021.113741

Han, Z.-H., & Zhang, K.-S. (2012). 17: Surrogate-based optimization. In O. Roeva (Ed.), Real-World Applications of Genetic Algorithms (pp. 343–362). InTech. https://doi.org/10.5772/36125

Henderi, H., Wahyuningsih, T., & Rahwanto, E. (2021). Comparison of Min-Max normalization and Z-Score Normalization in the K-nearest neighbor (kNN) Algorithm to Test the Accuracy of Types of Breast Cancer. International Journal of Informatics and Information Systems, 4(1), 13–20. https://doi.org/10.47738/ijiis.v4i1.73

Heras, N., de las, Roberts, E. P. L., Langton, R., & Hodgson, D. R. (2009). A review of metal separator plate materials suitable for automotive PEM fuel cells. Energy & Environmental Science, 2(2), 206–214. https://doi.org/10.1039/B813231N

Hodson, T. O. (2022). Root mean square error (RMSE) or mean absolute error (MAE): When to use them or not. Geoscientific Model Development Discussions, 15, 5481–5487. https://doi.org/10.5194/gmd-15-5481-2022

Karniadakis, G. E., Kevrekidis, I. G., Lu, L., Perdikaris, P., Wang, S., & Yang, L. (2021). Physics-informed machine learning. Nature Reviews Physics. https://doi.org/10.1038/s42254-021-00314-5

Lacki, P., Derlatka, A., Lacki, M., & Lachs, K. (2026). A Digital Twin for temperature prediction in the laser hardening process of NC10 steel. Computer Methods in Materials Science, 26(1), 5–22. https://doi.org/10.7494/cmms.2026.1.1023

Lin, K., Qiao, J., Shi, K., Dong, W., & Gu, D. (2023). Laser powder bed fusion of micro-channels for the application of proton exchange membrane fuel cell bipolar plates. CIRP Journal of Manufacturing Science and Technology, 43, 193–204. https://doi.org/10.1016/j.cirpj.2023.01.007

Modanloo, V., Jang, S., Lee, T., & Quagliato, L. (2025). Gradient Enhanced-Expert Informed Neural Network (GE-EINN) for forming depth prediction from a small-scale metal stamping dataset. Journal of Manufacturing Processes, 140, 224–240. https://doi.org/10.1016/j.jmapro.2025.02.052

Oliver, M. A., & Webster, R. (1990). Kriging: a method of interpolation for geographical information systems. International Journal of Geographical Information System, 4(3), 313–332. https://doi.org/10.1080/02693799008941549

Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378, 686–707. https://doi.org/10.1016/j.jcp.2018.10.045

Song, Y., Zhang, C., Ling, C. Y., Han, M., Yong, R.-Y., Sun, D., & Chen, J. (2020). Review on current research of materials, fabrication and application for bipolar plate in proton exchange membrane fuel cell. International Journal of Hydrogen Energy, 45(54), 29832–29847. https://doi.org/10.1016/j.ijhydene.2019.07.231

Winkler, R. L. (1994). Evaluating probabilities: Asymmetric scoring rules. Management Science, 40(11), 1395–1405. https://doi.org/10.1287/mnsc.40.11.1395

Xiao, W., Cai, H., Lu, W., Li, Y., Zheng, K., & Wu, Y. (2022). Multi-objective optimization with automatic simulation for partition temperature control in aluminum hot stamping process. Structural and Multidisciplinary Optimization, 65(3), 84. https://doi.org/10.1007/s00158-022-03190-4

Yan, Y., Wang, H., & Li, Q. (2015). The inverse parameter identification of Hill 48 yield criterion and its verification in press bending and roll forming process simulations. Journal of Manufacturing Processes, 20(1), 46–53. https://doi.org/10.1016/j.jmapro.2015.09.009

Yang, K.-T. (2008). Artificial neural networks (ANNs): a new paradigm for thermal science and engineering. Journal of Heat Transfer, 130(9). https://doi.org/10.1115/1.2944238

Zhang, H., Xu, C., Gao, T., Li, X., & Song, H. (2023). Identification of strain hardening behaviors in titanium alloys using tension tests and inverse finite element method. Journal of Mechanical Science and Technology, 37(7), 093001. https://doi.org/10.1007/s12206-023-0625-0

Zhang, R., Xu, Z., Peng, L., Lai, X., & Fu, M. W. (2021). Modelling of ultra-thin steel sheet in two-stage tensile deformation considering strain path change and grain size effect and application in multi-stage microforming. International Journal of Machine Tools and Manufacture, 164, 103713. https://doi.org/10.1016/j.ijmachtools.2021.103713

Zhong, Q., Hua, R., Wang, C., Cheng, L., Ma, Z., He, H., & Chen, F. (2023). Investigation on three-stage stamping of micro-channels with titanium ultra-thin sheet used for PEM fuel cell bipolar plates. The International Journal of Advanced Manufacturing Technology, 127(3), 1377–1389. https://doi.org/10.1007/s00170-023-11618-4