PSI - Issue 69

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

5

Figure 1: This figure presents the settling (Step 1 to 3) and spreading (Step 4 to 6) of the powder bed on the substrate as simulated using DEM in this study. It illustrates the resulting distribution of particles, which are arranged randomly for the layer under consideration for melt pool evaluation. For the purpose of this study, the thickness of this layer is set at 50 µm.

2.2. Simulation Setup and Material Properties In this study, the generation of NiTi powder particles within a powder bed was simulated. The subsequent step involved importing the file, representing the powder bed, into FLOW-3D for the formulation of the CFD model. Fig.2 visually represents the simulation domain, characterized by dimensions of 1000 micrometers in length, 300 micrometers in width, and 190 micrometers in height. The powder bed exhibited a thickness of 50 μm, alongside a substrate height of 90 μm. The computational domain was precisely meshed using a regular grid with a dimension of 3 μm, adopting a time step of 2e-8. This investigation implemented a split-Lagrangian approach to solve the volume of fluid and utilized an implicit technique for resolving the heat transfer field. On the other hand, surface tension was handled using an explicit method, and the GMERS (Generalized Minimal Residual) implicit technique was used to manage the pressure field. The material of choice for this study was NiTi, the properties of which are detailed in Table 1 and Fig.3. Time-variant material properties were predicted using Thermo-Calc (V-2022a), employing two distinct libraries, namely 'Scheil Solidification simulation' and 'property model calculation'.