An artificial neural network for surrogate modeling of stress fields in viscoplastic polycrystalline materials

The purpose of this work is the development of a trained artificial neural network for surrogate modeling of the mechanical response of elasto-viscoplastic grain microstructures. To this end, a U-Net-based convolutional neural network (CNN) is trained using results for the von Mises stress field from the numerical solution of initial-boundary-value problems (IBVPs) for mechanical equilibrium in such microstructures subject to quasi-static uniaxial extension. The resulting trained CNN (tCNN) accurately reproduces the von Mises stress field about 500 times faster than numerical solutions of the corresponding IBVP based on spectral methods. Application of the tCNN to test cases based on microstructure morphologies and boundary conditions not contained in the training dataset is also investigated and discussed.

Automated summary

The purpose of this work is to develop a trained artificial neural network for surrogate modeling of the mechanical response of elasto-viscoplastic grain microstructures. A U-Net-based convolutional neural network (CNN) is trained using numerical solutions for the von Mises stress field from initial-boundary-value problems (IBVPs) for mechanical equilibrium in such microstructures subject to quasi-static uniaxial extension. The resulting trained CNN (tCNN) accurately reproduces the von Mises stress field about 500 times faster than numerical solutions based on spectral methods. The application of the tCNN to test cases not contained in the training dataset is also investigated and discussed.