Efficient modeling of failure in composite materials with randomly distributed inclusions

This poster presents an efficient model for predicting failure in composite materials with randomly distributed inclusions. First, we conduct 2D analyses of stress and displacement fields using the finite element method, with particular emphasis on the material properties of the matrix and inclusions. By leveraging convolutional neural networks, we propose an efficient approach to accelerating these simulations.