This study presents a comprehensive review of thermal-fluid modeling techniques for welding processes based on computational fluid dynamics (CFD), with a particular focus on FLOW-3D and FLOW-3D WELD. Welding processes involve complex multiphysics phenomena including heat transfer, fluid flow, phase transformation, and vaporization, which significantly affect weld quality and defect formation. This review summarizes key numerical models such as free-surface tracking using the volume-of-fluid (VOF) method, heat source modeling for arc and laser welding, multiple reflection models for laser keyhole processes, and recoil pressure models. Furthermore, various welding processes?including conduction mode welding, keyhole laser welding, laser-arc hybrid welding, arc welding, beam oscillation welding, and dissimilar metal welding-are discussed to highlight the applicability of CFD-based simulations in predicting molten pool behavior, temperature distribution, and defect formation. Industrial applications in automotive, shipbuilding, and electronics industries are also reviewed to demonstrate the practical usefulness of these techniques. Despite significant advancements, several technical limitations remain, including challenges in multiphysics coupling, uncertainties in high-temperature material properties, and high computational cost. Future research directions are proposed, emphasizing the development of high-fidelity models, integration of multiphysics and multiscale approaches, incorporation of artificial intelligence for process optimization, and implementation of digital twin technologies. Overall, CFD-based welding simulation is expected to play a crucial role in advancing high-precision and high-efficiency manufacturing processes.
Cho et al. (Mon,) studied this question.