The application of artificial intelligence in mechanical simulation
DOI:
https://doi.org/10.61173/ga7rjg63Keywords:
artificial intelligence, mechanical simulation, solid mechanics, PINNAbstract
As computing capabilities surge forward and artificial intelligence progresses significantly, the integration of AI into mechanical simulation is expanding at a notable rate, substantially boosting advancements in engineering and scientific investigation realms. This article mainly explores three types of applications of artificial intelligence in mechanical simulation: surrogate models, physics-informed neural networks (PINNs), and adaptive grid refinement. These technologies improve the efficiency and accuracy of simulation and open up new research directions and application areas. In summary, the application of artificial intelligence technology in mechanical simulation has made significant progress. Still, it faces challenges such as strong data dependence, limited model generalization ability, and optimization of computational efficiency. Future research needs to conduct in-depth exploration into improving algorithm efficiency, enhancing model generalization ability, and expanding application scenarios. In addition, strengthening the combination of theoretical research and practical application will be an important direction for promoting the sustainable development of this field.References
[1] Jang, D.P., Fazily , P., & Yoon, J.W.(2021).Machine learningbased constitutive model for J2-plasticity.International Journal of Plasticity, 138, 102919.
[2] Hashemi, A., Jang, J., & Beheshti, J. (2023). A machine learning-based surrogate finite element model for estimating the dynamic response of mechanical systems. IEEE Access, p. 11, 54509–54525.
[3] Barmada, S., Fontana, N., Formisano, A., Thomopulos, D., & Tucci, M. (2021). A deep learning surrogate model for topology optimization. IEEE Transactions on Magnetics, 57(6), 1-4.
[4] Xu, Y., Weng, H., Ju, X., Ruan, H., Chen, J., Nan, C., ... & Liang, L. (2021). A method for predicting mechanical properties of composite microstructure with a reduced dataset based on transfer learning. Composite Structures, 275, 114444.
[5] Weng, H., & Yuan, H. (2023). Quantitative representation of directional microstructures of single-crystal superalloys in cyclic crystal plasticity based on neural networks. International Journal of Plasticity, 170, 103757.
[6] Heider, Y., Wang, K., & Sun, W. (2020). SO (3)-invariance of informed-graph-based deep neural network for anisotropic elastoplastic materials. Computer Methods in Applied Mechanics and Engineering, 363, 112875.
[7] Schiassi, E., D’Ambrosio, A., Drozd, K., Curti, F., & Furfaro, R. (2022). Physics-informed neural networks for optimal planar orbit transfers. Journal of Spacecraft and Rockets, 59(3), 834- 849.
[8] Huang, K., Krü gener , M., Brown, A., Menhorn, F., Bungartz , H.J., & Hartmann, D.(2021).Machine learning-based optimal mesh generation in computational fluid dynamics.arXiv preprint arXiv:2102.12923.
Downloads
Published
Issue
Section
License
Copyright (c) 2024 by the authors.
This work is licensed under a Creative Commons Attribution 4.0 International License.
