A novel thermo-geometrical model for accurate keyhole porosity prediction in Laser Powder-Bed Fusion

Abstract

AbstractWhen performing Laser Powder-Bed Fusion, undesired physical phenomena, such as balling, preballing and keyhole, must be avoided in order to achieve high-quality products. To date, keyhole-free process parameters can be identified either using demanding empirical methods or complex numerical simulations, while only a few analytical models can be found in literature for this purpose. In this work, state-of-the-art analytical models for predicting keyhole porosity were summarized and proved to be rather inaccurate because they are only based on thermodynamic principles, whereas they neglect the geometry and both the kinetics and kinematics of the keyhole cavity, which do also influence cavity collapse and porosity formation. Here an innovative physics-based semi-analytical model for predicting the formation of keyhole-related porosities was conceived, in which both thermodynamic and geometrical factors are taken into account. The proposed model was validated by performing single tracks experiments on Ti6Al4V according to a full factorial DoE on laser power and scanning speed. Produced samples were cross-sectioned and analyzed to evaluate keyhole porosity formation. The comparison between experimental data and model predictions confirmed the higher accuracy of the new model with respect to state of the art models. Besides improving the understanding of the keyhole phenomenon, the proposed model may provide a novel, effective and simple tool for fast process parameter optimization in industry.