Additive manufacturing (AM) opens the possibility to make lighter and more complex geometries using less material and shortens the time from idea to prototype. This requires a better understanding of the atomization process to produce high quality metallic powder as a feedstock. The atomization process requires a better understanding how the gas and melt interact to optimize the breakup of the metal stream. Computational fluid dynamics (CFD) is used to visualize fluid flow phenomena to create a model representation of the supersonic gas in the atomisation process in a close-coupled atomizer. Results show the transient behavior of the gas flow including turbulent fluctuations and typical flow structures such as shock waves, Mach discs, local density, and pressure changes, etc. The model developed is validated with Schlieren imaging and frequency analysis (FFT). A parametric study on the turbulence models available, mesh density, time stepping, sub-grid models as well as geometry assumptions such as 2D vs 3D ¼ section vs full 3D is carried out to determine the appropriate parameters for the simulations. The final aim is to empower the powder producers with a fast and reliable model which allows testing of geometry, pressure, and temperature variations for optimization of the atomization process.