Modeling method for predicting part-scale geometric morphology considering manufacturing system states of laser-based directed energy deposition

Abstract

Laser-based directed energy deposition (DED-LB) represents a promising additive manufacturing technique to produce large-scale parts and an urgent requirement of geometric accuracy with high-value products. However, achieving high forming accuracy remains a challenge, primarily due to extensive back-and-forth toolpaths and layer-by-layer thermal accumulation. During the actual manufacturing process, planned toolpaths dominate the dynamic spatiotemporal input of the laser beam and powder stream, which involves intricate heat transfer within deposition layers and results in potential geometric deviations. This study developed a virtual simulation framework that couples the manufacturing system states with a heat transfer model to predict the part-scale geometric morphology of deposited parts. An identification method was proposed to obtain kinematic timing sequence data from the dynamic manufacturing system with a prediction error of less than 1%, which considers the acceleration-deceleration phenomenon and laser switch time delays. The constructed model clearly elucidates heat dissipation mechanisms and visually illustrates the diverse deposition features caused by differences in melt pool area under different path patterns. Furthermore, the findings strongly confirm that the manufacturing system states significantly influence the geometric morphology, heat-affected depth, and heat-affected steady-state temperature. Simplified simulations that ignore the actual system states may overestimate the volume of deposited parts by a maximum of 53 % due to excessive thermal accumulation. The proposed virtual simulation method offers essential insights for practical guidance in path planning and machine tool configuration during the DED-LB process.