PSI - Issue 69

Mohammadjavad Abdollahzadeh et al. / Procedia Structural Integrity 69 (2025) 2–19

11

and radiation, emphasizing their importance in comprehensively addressing and modeling the LPBF process computationally. Furthermore, the quantity of vaporized mass warrants examination. In essence, the process of evaporation is a critical element that must be factored into the development of a numerical solution.

Fig.6: Energy interaction in LPBF process. a) Simulation of the melt Pool, b) Schematic of the heat transfer and laser beam distribution.

3. Result and Discussion 3.1. Numerical Modeling Validation

Melt-pool fluid dynamics play a crucial role in validating simulation results against experimental data. A major challenge in this process is the limited availability of complete material properties for NiTi. While some properties were obtained using Thermo-Calc and available literature, not all experimental interactions can be fully captured, which may contribute to discrepancies in melt-pool behavior. Additionally, mesh resolution significantly affects simulation accuracy—coarser meshes tend to oversimplify physical phenomena, while finer meshes increase computational cost and can introduce numerical instability. To ensure both accuracy and efficiency, a mesh independence study was performed. Preliminary simulations with coarser meshes showed noticeable deviations in predicted melt-pool dimensions. However, further refinement below 3 μm resulted in minimal improvement, confirming that a 3 μm mesh strikes an optimal balance between fidelity and computational efficiency. To validate the numerical model under comparable processing conditions, the study by Chernyshikhin et al. [51] was selected. This work systematically investigated NiTi melt pool formation using parameters closely aligned with those used in this research and provides detailed experimental data on melt pool geometry. The same alloy system (NiTi) and overlapping process conditions make it a strong reference for benchmarking. Figure 7 presents a comparative analysis between the melt-pool morphology predicted in this study and the experimental findings by Chernyshikhin et al., showing strong agreement with errors below 2.8% for height and 3.1% for width. Additionally, variations in melt pool depth and width with increasing laser power are illustrated in Figure 8. The comparison between the numerical and experimental results demonstrates reasonable correlation, with an overall error of approximately 15%.