Extraction of mechanical properties of materials through deep learning from instrumented indentation
The data and code for the paper L. Lu, M. Dao, P. Kumar, U. Ramamurty, G. E. Karniadakis, & S. Suresh. Extraction of mechanical properties of materials through deep learning from instrumented indentation. Proceedings of the National Academy of Sciences, 117(13), 7052-7062, 2020.
All the data is in the folder data.
All the code is in the folder src. The code depends on the deep learning package DeepXDE v1.1.2. If you use DeepXDE>1.1.2, you need to set standardize=True
in dde.data.MfDataSet()
.
- data.py: The classes are used to read the data file. Remember to uncomment certain line in
ExpData
to scaledP/dh
. - nn.py: The main functions of multi-fidelity neural networks.
- model.py: The fitting function method. Some parameters are hard-coded in the code, and you should modify them for different cases.
- fit_n.py: Fit strain-hardening exponent.
- mfgp.py: Multi-fidelity Gaussian process regression.
If you use this code for academic research, you are encouraged to cite the following paper:
@article{Lu7052,
author = {Lu, Lu and Dao, Ming and Kumar, Punit and Ramamurty, Upadrasta and Karniadakis, George Em and Suresh, Subra},
title = {Extraction of mechanical properties of materials through deep learning from instrumented indentation},
volume = {117},
number = {13},
pages = {7052--7062},
year = {2020},
doi = {10.1073/pnas.1922210117},
journal = {Proceedings of the National Academy of Sciences}
}
To get help on how to use the data or code, simply open an issue in the GitHub "Issues" section.