A physics-constrained neural network for crystal plasticity modelling of FCC materials

Article Materials Science, Multidisciplinary

An artificial neural network-based model for roping prediction in aluminum alloy sheet

Yuanzhe Hu et al.

Summary: An artificial neural network-based (ANN) model is developed to predict thickness strain and corresponding surface roping in aluminum alloy. Combined with the crystal plasticity finite element model (CPFEM), an artificial orientation map generation algorithm is proposed to produce reliable datasets for training and validating the ANN model. The ANN-based model demonstrates good prediction capability and generality in multiple validation cases.

ACTA MATERIALIA (2023)

Add to Collection

Article Multidisciplinary Sciences

Analyses of internal structures and defects in materials using physics-informed neural networks

Enrui Zhang et al.

Summary: In this study, a general framework based on physics-informed neural networks is proposed for identifying unknown geometric and material parameters in materials' internal structures and defects. By using a mesh-free method, the geometry of the material is parameterized, and the effectiveness of the method is validated using constitutive models. The framework can be applied to other inverse problems involving unknown material properties and deformable geometries.

SCIENCE ADVANCES (2022)

Add to Collection

Article Engineering, Mechanical

From CP-FFT to CP-RNN: Recurrent neural network surrogate model of crystal plasticity

Colin Bonatti et al.

Summary: In this study, RNN models were used to investigate the mechanical behavior of aluminum alloys and reasonable approximations were obtained. A methodology to reduce numerical instabilities in the finite element implementation was also proposed.

INTERNATIONAL JOURNAL OF PLASTICITY (2022)

Add to Collection

Article Nanoscience & Nanotechnology

Machine learning based surrogate modeling approach for mapping crystal deformation in three dimensions

Anup Pandey et al.

Summary: The machine learning based surrogate modeling method presented in this study can predict the spatially resolved 3D crystal orientation evolution of polycrystalline materials under uniaxial tensile loading. It is significantly faster than existing crystal plasticity methods and enables simulation of large volumes that would be otherwise computationally prohibitive. This approach surpasses existing ML-based modeling results by providing local 3D full-field predictions rather than just average values or being limited to 2D structures.

SCRIPTA MATERIALIA (2021)

Add to Collection

Article Engineering, Mechanical

A new ANN based crystal plasticity model for FCC materials and its application to non-monotonic strain paths

Olga Ibragimova et al.

Summary: Machine learning methods are commonly used for pattern recognition and can offer substantial computational advantages over conventional numerical methods. The proposed framework in this research, combining machine learning and crystal plasticity, accurately predicts stress-strain behavior and texture evolution for a variety of materials within the FCC family.

INTERNATIONAL JOURNAL OF PLASTICITY (2021)

Add to Collection

Review Physics, Applied

Physics-informed machine learning

George Em Karniadakis et al.

Summary: Physics-informed learning seamlessly integrates data and mathematical models through neural networks or kernel-based regression networks for accurate inference of realistic and high-dimensional multiphysics problems. Challenges remain in incorporating noisy data seamlessly, complex mesh generation, and addressing high-dimensional problems.

NATURE REVIEWS PHYSICS (2021)

Add to Collection

Article Engineering, Mechanical