Review: finite element modeling, artificial intelligence, digital twins and applications for additive manufacturing of refractory alloys

Abstract

Refractory alloys (RAs), such as tungsten-, molybdenum-, tantalum-, and niobium-based systems, and the emerging refractory high-entropy alloys (RHEAs), are gaining more and more attention for extreme-environment applications due to their exceptional high-temperature strength, thermal stability, creep resistance, and radiation tolerance. However, their fabrication still remains a challenge due to their high melting temperatures, oxidation sensitivity, ductile-to-brittle transition behavior and susceptibility to cracking and residual stress formation. Additive manufacturing (AM) has emerged as a promising route to fabricate complex refractory alloy components with improved material utilization and design flexibility. However, the steep thermal gradients, fast solidification rates and complex thermo-metallurgical phenomena associated with AM often lead to microstructural heterogeneity, porosity, distortion, and performance variability. The present review aims to provide a comprehensive overview of the synergistic combination of finite element modeling (FEM) and artificial intelligence (AI) in the additive manufacturing of refractory alloys. The capabilities of the main AM technologies, such as laser powder bed fusion, laser directed energy deposition, electron beam powder bed fusion, and wire arc additive manufacturing, are critically reviewed. Recent advances in thermal, mechanical, microstructural, and multiscale FEM approaches are discussed with emphasis on process–structure–property relationships. Furthermore, a systematic review on the implementation of machine learning, deep learning, physics-informed neural networks and digital twins for process optimization, defect prediction and quality assurance is provided. The synergistic integration of AI and FEM is highlighted as a key pathway for predictive and autonomous manufacturing. Finally, current challenges, knowledge gaps, and future research opportunities are identified to accelerate the development of intelligent manufacturing frameworks for next-generation refractory alloy components.