journal article·2022-04Open access

Learning constitutive relations of plasticity using neural networks and full-field data

Yin Zhang, Qing-Jie Li, Ting Zhu, Ju Li

Abstract

Neural networks (NNs) have demonstrated strong capabilities of learning constitutive relations from big data. However, most NN-based constitutive models require experimental data from a considerable number of stress–strain paths that are expensive to collect. Here, we develop a hybrid finite element method - NN (FEM-NN) framework for learning the constitutive relations from full-field data. As a result, the non-uniform displacement field from a deformed sample with geometrical inhomogeneities can be used for training NNs. Such full-field data have the advantage of providing many different stress–strain paths at different locations in the sample by a single test, thereby enabling the highly efficient training of NNs. We apply FEM-NN simulations to learn the constitutive relations of several model materials characterized by rate-independent J 2 plasticity. These FEM-NN studies demonstrate that the trained NNs produce the constitutive relations of plasticity with high accuracy and efficiency.
Source: expertdata · Quality: candidate unverified · Source record ↗

Research context

OpenAlex topics: Force Microscopy Techniques and Applications · Elasticity and Material Modeling · Microstructure and mechanical properties

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

References (38)

Show source-provided references
1. Constitutive analysis of elastic–plastic crystals at arbitrary strain10.1016/0022-5096(72)90017-8 ↗
2. Hill · 1998
3. Argon · 2008
4. Extraction of mechanical properties of materials through deep learning from instrumented indentation10.1073/pnas.1922210117 ↗
5. PPINN: Parareal physics-informed neural network for time-dependent PDEs10.1016/j.cma.2020.113250 ↗
6. Knowledge extraction and transfer in data-driven fracture mechanics10.1073/pnas.2104765118 ↗
7. Deep neural network method for predicting the mechanical properties of composites10.1063/1.5124529 ↗
8. Deep learning method for determining the surface elastic moduli of microstructured solids10.1016/j.eml.2021.101226 ↗
9. Implicit constitutive modelling for viscoplasticity using neural networks10.1002/(SICI)1097-0207(19980930)43:2<195::AID-NME418>3.0.CO;2-6 ↗
10. Prediction of nonlinear viscoelastic behavior of polymeric composites using an artificial neural network10.1016/j.ijplas.2005.09.002 ↗
11. Exploring elastoplastic constitutive law of microstructured materials through artificial neural network—a mechanistic-based data-driven approach10.1115/1.4047208 ↗
12. Machine-learning based temperature- and rate-dependent plasticity model: Application to analysis of fracture experiments on DP steel10.1016/j.ijplas.2019.02.012 ↗
13. Strain rate and temperature dependent fracture of aluminum alloy 7075: Experiments and neural network modeling10.1016/j.ijplas.2020.102788 ↗
14. Physics-driven machine learning model on temperature and time-dependent deformation in lithium metal and its finite element implementation10.1016/j.jmps.2021.104481 ↗
15. On the potential of recurrent neural networks for modeling path dependent plasticity10.1016/j.jmps.2020.103972 ↗
16. Recurrent neural network modeling of the large deformation of lithium-ion battery cells10.1016/j.ijplas.2021.103072 ↗
17. A recurrent neural network-accelerated multi-scale model for elasto-plastic heterogeneous materials subjected to random cyclic and non-proportional loading paths10.1016/j.cma.2020.113234 ↗
18. Using neural networks to represent von mises plasticity with isotropic hardening10.1016/j.ijplas.2020.102732 ↗
19. Learning constitutive relations from indirect observations using deep neural networks10.1016/j.jcp.2020.109491 ↗
20. Learning constitutive relations using symmetric positive definite neural networks10.1016/j.jcp.2020.110072 ↗
21. Sobolev training of thermodynamic-informed neural networks for interpretable elasto-plasticity models with level set hardening10.1016/j.cma.2021.113695 ↗
22. Application of artificial neural networks in micromechanics for polycrystalline metals10.1016/j.ijplas.2019.05.001 ↗
23. High-entropy alloys10.1038/s41578-019-0121-4 ↗
24. Tuning element distribution, structure and properties by composition in high-entropy alloys10.1038/s41586-019-1617-1 ↗
25. Hybrid FEM-NN models: Combining artificial neural networks with the finite element method10.1016/j.jcp.2021.110651 ↗
26. The FEniCS project version 1.5
27. dolfin-adjoint 2018.1: automated adjoints for FEniCS and firedrake10.21105/joss.01292 ↗
28. On the limited memory BFGS method for large scale optimization10.1007/BF01589116 ↗
29. Applications of digital-image-correlation techniques to experimental mechanics10.1007/BF02325092 ↗
30. 2013
31. Unified form language: A domain-specific language for weak formulations of partial differential equations10.1145/2566630 ↗
32. Consistent tangent operators for rate-independent elastoplasticity10.1016/0045-7825(85)90070-2 ↗
33. Goodfellow · 2016
34. SciPy 1.0: fundamental algorithms for scientific computing in Python10.1038/s41592-019-0686-2 ↗
35. N. Schlömer, meshio: Tools for mesh files, https://github.com/nschloe/meshio.
36. Clevert · 2015
37. Ayachit · 2015
38. Machine learning in nuclear materials research10.1016/j.cossms.2021.100975 ↗