Convolutional Neural Network for Enhancement of Localization in Granular Representative Unit Cells

Abstract

A convolutional neural network was used to enhance the localization of strain and stress for a generalized method of cells model of a metallic microstructure. Enhanced shear strains, measured in terms of the linear regression coefficients as a function of ground truth strains, were improved from inaccurate and uncorrelated ([Formula: see text], [Formula: see text]) to accurate and well correlated ([Formula: see text], [Formula: see text]) relative to ground truth ([Formula: see text], [Formula: see text]). Kernel sizes of 2 or 3 were effective in the convolutional neural network (padding = “same”). graphical processing unit (GPU)-parallelized enhancement costs were low after training (range 0.41–3.45%) compared to the baseline generalized method of cells, and are significantly faster than finite element. The accuracy of enhanced localized shear strains and stress is expected to yield benefits for damage progression models, especially in the context of hierarchical multiscale methods where the generalized method of cells is applied at the intermediate scale.