WaiChing Sun

1 Introduction Advanced manufacturing techniques allow for the creation of materials with properties tailored to their use. Multiscale modeling can accurately capture the behavior of these materials, yet its use is limited by the high computational cost of simulations. The need for more computationally efficient models, combined with the progress in machine learning, has resulted in numerous data-driven models [1-6].

Abstract We present a data-driven framework for the multiscale modeling of anisotropic finite strain elasticity based on physics-augmented neural networks (PANNs). Our approach allows the efficient simulation of materials with complex underlying microstructures which reveal an overall anisotropic and nonlinear behavior on the macroscale.

1 INTRODUCTION Sand is the second most consumed natural resource material in the world1 (the first being water). Every year, 50 billion tonnes of sand are used as raw material for concrete, glass, and electronics.