Bed-based additive manufacturing (AM) techniques, in which thin layers of metal powder are locally fused together by a laser, allow for the production of components with complex designs. While the metal AM technology is relatively matured, flexible and accurate production of high quality metal powder is less developed. Gas atomization is a production technique for metal powder, in which high pressure gas jets rupture a molten metal stream into fine droplets, which subsequently solidify into a powder. A small scale gas atomization process for the production of high-quality metal powder can greatly enhance the versatility and sustainability of powder bed AM techniques.
The gas atomization process is typically described in terms of a primary and secondary breakup stage. This work focuses on the modelling of the primary breakup of the metal melt, covering the initial breakup into large droplets and ligaments. The secondary atomization describes the further breakup of these droplets into fine elements. High local Weber and Reynolds numbers are encountered as a result of supersonic gas jets impinging on the melt stream. Large velocity and temperature gradients increase the computational complexity of the breakup process.
While in previous studies mainly the gas dynamics were studied, the current modelling strategy is aimed at understanding the primary breakup process for both the melt and gas phase. To this end, the gas-melt interface is modelled using a geometric Volume of Fluid (VOF) method for compressible flow in the open-source OpenFOAM software.
Large Eddy Simulations (LES) are applied, in order to partially resolve the local turbulent flow structures. Special attention is devoted to the local resolution of the computational grid, with respect to requirements of both the VOF and LES. Adaptive mesh refinement is applied at the phase interface to ensure sufficient resolution of the VOF method. The numerical methods are validated separately prior to application in primary atomization.
A parameter study is performed based on design of experiments principles. Statistics of the droplets are collected at several locations in the spray, in order to analyze the impact of process parameters on the primary breakup process. The gathered droplet statistics allow to separately model the secondary breakup process using less computationally intensive modelling techniques in the future.