Machine learning enabled identification of sheet metal localization

Abstract

The Forming Limit Curve (FLC), which describes the maximum applicable strain before localization, depends on the particular material, but also on the applied load and history of the load. Recent investigations have shown that the non-proportional loading effect on the FLC can be predicted with data-driven or machine-learning-based methods. Here we compare different ML methods to their applicability in predicting localization points under multi-segmented non-proportional loading. Therefore, an FE-based metamodel is developed that allows imposing an arbitrary loading history on sheet metal to predict the point of localization. A series of virtual experiments are conducted with this metamodel to generate a database of bi-linear loading paths that are used for training. Different ML-based methods were used to predict the localization point based on the strain history data. The 1D-Convolutional Neural Network (1D-CNN), with the ability to learn dependency between input features, has the best accuracy in predicting the localization point.