Abstract

Recent progress in additive manufacturing has brought renewed focus on the use of Directed Energy Deposition (DED) for fabricating complex components from superalloys, particularly in aerospace and power generation applications. One of the central challenges in DED lies in accurately predicting microstructural evolution during the rapid melting and solidification processes inherent to the technique. This abstract presents recent advances in predicting microstructural evolution in superalloys using DED process data, with an emphasis on the integration of experimental insights, computational modeling, and data-driven methods. Advanced monitoring techniques such as in-situ thermal imaging, melt pool sensors, and high-speed cameras have provided detailed temporal and spatial datasets that reflect process-induced thermal histories. These datasets are critical for understanding solidification kinetics, grain morphology, dendritic arm spacing, and phase transformations in real-time. Coupling these datasets with finite element analysis (FEA) and cellular automata (CA) models has enabled high-fidelity simulation of grain growth and texture evolution under varying DED conditions. Machine learning models are now being trained on DED data to predict microstructural features such as grain orientation, residual stresses, and porosity with increasing accuracy. By leveraging big data from in-situ sensors and post-process characterizations (e.g., EBSD, XRD, and SEM), these models offer fast, predictive capabilities that complement traditional physics-based approaches. Moreover, the development of digital twin frameworks for DED processes is accelerating the shift toward predictive and adaptive manufacturing, where real-time feedback and microstructural control are achievable. Significant progress has also been made in linking process parameters (e.g., laser power, scanning speed, layer thickness) to resultant microstructures and mechanical properties. These correlations are essential for qualification and certification of DED-produced superalloy components, where consistency, durability, and performance under extreme conditions are paramount. This paper highlights the confluence of experimental techniques, physics-based models, and AI tools as a transformative approach to predicting and controlling microstructural evolution in DED-processed superalloys.