Artificial neural networks as material models for finite element analysis

Abstract

An obstacle for development of structural components is often the adequate description of materials for computational methods, especially in finite element analysis. The substitution of classical constitutive material models with data-driven models using machine learning techniques could be a new strategy, providing a faster description, faster computation and efficient parameter identification for the modelling of materials. For this purpose, an artificial neural network (ANN) is trained, which learns the relationship between strains and stresses and is capable of modelling the hypo-elastic behaviour from a constitutive model. To train the ANN, artificial stress-strain curves are generated using a programming interface for LS-DYNA’s material model driver. A hyperparameter study finds the optimal set of hyperparameters for training of the neural network. Finally, the trained ANN is implemented in a user-defined subroutine, which provides a link between modern machine learning frameworks and commercial FEA codes. Explicit simulations using the ANN model run stable and show good agreement in global response as well as on local element stress level.