Multiphysics Simulation of Wire-Laser Directed Energy Deposition of NiTi: Melt Pool Dynamics and Experimental Validation
摘要
Directed energy deposition using a laser beam and wire feedstock (DED-LB/w) is a promising route for fabricating NiTi components, yet its usable parameter space is narrow because melt pool dynamics, evaporation, and thermal gradients can modify composition and functional behavior. A three-dimensional multiphysics CFD model is developed to simulate melt pool evolution during DED-LB/w of NiTi. The novelty of this work lies in developing and experimentally validating, to the authors’ knowledge, among the first DED-LB/w-specific multiphysics CFD frameworks for NiTi that simultaneously resolves free-surface evolution, Marangoni-driven flow, buoyancy, evaporation-induced heat loss, and recoil pressure while quantitatively linking these transient melt pool physics to experimentally measured single-track geometries. The model solves the coupled mass, momentum, and energy equations using a VOF free-surface formulation with enthalpy–porosity solidification, temperature-dependent properties, Marangoni convection, buoyancy, evaporation, and recoil pressure. Simulations were performed over 600–1200 W and 300–700 mm/min, and validation was conducted using experimentally measured melt pool depth, width, and buildup height from transverse cross-sections. The 600 and 800 W validation cases are presented in the main text, while the 1000 and 1200 W cases are included in the Supplementary Material. Increasing power shifts the melt pool from compact, conduction-dominated behavior to larger, convection-dominated pools with stronger unsteadiness and evaporation, whereas increasing scan speed reduces thermal exposure and raises the risk of insufficient fusion. Beyond validating melt pool dimensions, the study introduces a process physics-based interpretation of DED-LB/w NiTi through temperature-history analysis, heat flux–temperature correlations, and correlation heatmaps of thermophysical and flow-related variables, providing a more mechanistic basis for process window selection than geometry comparison alone. The results suggest a practical operating window at 600–800 W and 500–600 mm/min that balances adequate penetration with reduced overheating and provide a foundation for coupling DED-LB/w thermal histories with microstructure and phase transformation models for functional-property prediction.