Analysis of Power-Law Fin-Type Problems using Physics Informed Neural Networks

Göçer, M.; Coşkun, S.B.; Atay, M.T.

doi:10.2478/jaes-2025-0026

Abstract

This study aims to model the temperature distribution in a single fin subjected to steady one-dimensional heat conduction with nonlinear thermal behavior. For the modeling and solution of the problem, the Physics-Informed Neural Networks (PINNs) architecture was used. The temperature-dependent heat conduction problem and the nonlinear boundary conditions of this problem were formulated with a differential equation. With the help of the PINN architecture, the loss function was minimized in order to reduce the difference between the true value and the predicted value. During this minimization process, the PINN architecture was forced to be consistent with the physical laws. The results obtained after training the PINN architecture exhibit successful performance in terms of accuracy and reliability when compared with the results in the literature. These findings highlight the potential of PINNs as a powerful alternative to conventional methods for solving complex nonlinear heat conduction problems.

References

Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard, M., and Kudlur, M., 2016, Tensorflow: A System for Large-Scale Machine Learning. OSDI’16: Proceedings of the 12th USENIX conference on Operating Systems Design and Implementation, 16, pp. 265–283. doi: 10.48550/arXiv.1605.08695.
Search in Google Scholar Back to article
Atay, M.T., Coşkun, S.B., 2008. Comparative Analysis of Power-Law Fin-Type Problems Using Variational Iteration Method and Finite Element Method. Math. Probl. Eng., 2008, 635231. doi:10.1155/2008/635231.
Search in Google Scholar Back to article
Bai, J., Rabczuk, T., Gupta, A., Alzubaidi, L., Gu, Y., 2023. A physics-informed neural network technique based on a modified loss function for computational 2D and 3D solid mechanics. Comput. Mech., 71, pp. 543-562. doi: 10.1007/s00466-022-02252-0.
Search in Google Scholar Back to article
Cai, S., Wang, Z., Wang, S., Perdikaris, P., Karniadakis, G.E., 2021. Physics-Informed Neural Networks for Heat Transfer Problems. J. Heat Transf., 143, 060801. doi: 10.1115/1.4050542.
Search in Google Scholar Back to article
Chen, Y., Lu, L., Karniadakis, G.E., Dal Negro, L., 2020. Physics-informed neural networks for inverse problems in nano-optics and metamaterials. Opt. Express, 28, 11618. doi: 10.1364/OE.384875.
Search in Google Scholar Back to article
Fang, Z., Zhan, J., 2020. Deep physical informed neural networks for metamaterial design. IEEE Access, 8, pp. 24506–24513. doi: 10.1109/ACCESS.2019.2963375.
Search in Google Scholar Back to article
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. Comput. Methods Appl. Mech. Engrg., 379, 113741. doi: 10.1016/j.cma.2021.113741.
Search in Google Scholar Back to article
Hao, W.Q., Tan, L., Yang, X.G., Shi, D.Q., Wang, M.L., Miao, G.L., Fan, Y.S., 2023. A physics-informed machine learning approach for notch fatigue evaluation of alloys used in aerospace. Int. J. Fatigue, 170, 107536. doi: 10.1016/j.ijfatigue.2023.107536.
Search in Google Scholar Back to article
Hu, Z., Jagtap, A.D., Karniadakis, G.E., Kawaguchi, K., 2019. When Do Extended Physics-Informed Neural Networks (XPINNs) Improve Generalization? arXiv, arXiv:2109.09444. doi: 10.1137/21M1447039.
Search in Google Scholar Back to article
Kern, D.Q., Kraus, A.D., 1972. Extended Surface Heat Transfer. McGraw-Hill, New York, NY, USA. ISBN: 978-0070341951.
Search in Google Scholar Back to article
Liaw, S.P., Yeh, R.H., 1994. Fins with temperature dependent surface heat flux—I: single heat transfer mode. Int. J. Heat Mass Transf., 37(10), pp. 1509–1515. doi: 10.1016/0017-9310(94)90152-X.
Search in Google Scholar Back to article
Madir, B.E., Luddens, F., Lothode, C., Danaila, I., 2024. Physics Informed Neural Networks for heat conduction with phase change. arXiv, arXiv:2410.14216. doi: 10.48550/arXiv.2410.14216.
Search in Google Scholar Back to article
Ning, L., Cai, Z., Dong, H., Liu, Y., Wang, W., 2023. A peridynamic-informed neural network for continuum elastic displacement characterization. Comput. Methods Appl. Mech. Engrg., 407, 115909. doi: 10.1016/j.cma.2023.115909.
Search in Google Scholar Back to article
Paszke, A., Gross, S., Chintala, S., Chanan, G., Yang, E., DeVito, Z., Lin, Z., Desmaison, A., Antiga, L., and Lerer, A., 2017. Automatic Differentiation in Pytorch. 31st Conference on Neural Information Processing Systems (NIPS 2017), CA, USA. https://openreview.net/forum?id=BJJsrmfCZ.
Search in Google Scholar Back to article
Raissi, M., Perdikaris, P., Karniadakis, G.E., 2019a. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys., 378, pp. 686–707. doi: 10.1016/j.jcp.2018.10.045.
Search in Google Scholar Back to article
Raissi, M., Wang, Z., Triantafyllou, M.S., Karniadakis, G.E., 2019b. Deep learning of vortex-induced vibrations. J. Fluid Mech., 861, pp. 119–137. doi: 10.48550/arXiv.1808.08952.
Search in Google Scholar Back to article
Raissi, M., Yazdani, A., Karniadakis, G.E., 2020. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. Science, 367, pp. 1026–1030. doi: 10.1126/science.aaw4741.
Search in Google Scholar Back to article
Russel, C., 2023. Applying Physics-Informed Neural Networks to the Solution of Parametric, Nonlinear Heat Conduction Problems. M.Sc. Thesis, University of Michigan-Flint, University of Michigan Library, Michigan, USA. doi: 10.7302/21927.
Search in Google Scholar Back to article
Sen, A.K., Trinh, S., 1986. An exact solution for the rate of heat transfer from a rectangular fin governed by a power law-type temperature dependence. J. Heat Transfer, 108(2), pp. 457-459. doi: 10.1115/1.3246946.
Search in Google Scholar Back to article
Shukla, K., Jagtap, A.D., Karniadakis, G.E., 2021. Parallel physics-informed neural networks via domain decomposition. J. Comput. Phys., 447, 110683. doi: 10.48550/arXiv.2104.10013.
Search in Google Scholar Back to article
Zhu, Q., Zhao, Z., Yan, J., 2023. Physics-informed machine learning for surrogate modeling of wind pressure and optimization of pressure sensor placement. Comput. Mech., 71, pp. 481-491. doi: 10.1007/s00466-022-02251-1.
Search in Google Scholar Back to article

Analysis of Power-Law Fin-Type Problems using Physics Informed Neural Networks

Abstract

Paradigm

My account