A physics-guided deep generative model for predicting melt pool behavior in laser powder bed fusion additive manufacturing

Abstract

Abstract Laser powder bed fusion (LPBF) is a promising metal additive manufacturing process that enables the production of highly intricate geometries. Achieving consistent quality and repeatability in LPBF lies in accurately predicting and controlling melt pool behavior. Recent studies have primarily utilized data-driven approaches using real-time melt pool monitoring (MPM) data. However, these methods often lack accuracy and interpretability, primarily because they rely on data without adequately considering the underlying physical mechanisms related to melt pool formation. To address this issue, our study introduces a novel physics-guided deep generative model to predict melt pool behavior in LPBF. We employ a Convolutional Neural Network Transformer Generative Adversarial Network to predict future MPM images, leveraging a physics-based model to enhance the accuracy and interpretation of our predictions. Our experimental validation highlights the model’s effectiveness and accuracy in predicting melt pool behaviors in LPBF. A comparison with related studies shows that the proposed model achieves better prediction accuracy, demonstrating improvements in melt pool geometry and image quality. This advancement in melt pool modeling significantly contributes to the LPBF, promising to improve its process control and part quality.