Physics-informed neural network-based computational solid mechanics

This repository provides numerical examples of physics-informed neural network-based computational solid mechanics framework.

Physics-informed neural network (PINN) has recently gained increasing interest in computational mechanics.

This work aims to help the researchers who are interested in the PINN-based solid mechanics solver to have a clear insight into this emerging area.

Numerical examples include:

• 1D stretching plate problem (implemented by collocation and energy-based loss functions)
• 2D in-plain stretching plate problem (implemented by collocation loss function)
• 2D in-plain stretching plate problem (implemented by energy-based loss function)
• 3D streching cube problem (implemented by collocation loss function)

Information regarding the numerical examples, please refer to our paper.

Enviornmental settings

• TensorFlow 2.8.0
• Keras 2.8.0
• SciPy 1.8.0
• Matplotlib 3.6.0

Paper link

Now, the paper has been accepted by the International Journal of Computational Methods(IJCM):
https://doi.org/10.1142/S0219876223500135
Besides, the preprint version can be found at:
https://arxiv.org/abs/2210.09060

Cite as

[1] J. Bai, H. Jeong, C.P. Batuwatta-Gamage, S. Xiao, Q. Wang, C.M. Rathnayaka, L. Alzubaidi, G.-R. Liu, Y. Gu, An introduction to programming Physics-Informed Neural Network-based computational solid mechanics, International Journal of Computational Methods, (2023), DOI: 10.1142/S0219876223500135.
or
[1] J. Bai, H. Jeong, C.P. Batuwatta-Gamage, S. Xiao, Q. Wang, C.M. Rathnayaka, L. Alzubaidi, G.-R. Liu, Y. Gu, An introduction to programming Physics-Informed Neural Network-based computational solid mechanics, Arxiv preprint arXiv:2210.09060, (2022).

Contact us

For questions regarding the code, please contact:

jinshuai.bai@hdr.qut.edu.au

yuantong.gu@qut.edu.au