An Experimental Approach for Characterizing Gas Flow in the Build Chamber of Laser Powder Bed Fusion Systems Utilizing Particle Image Velocimetry

Abstract

Laser Powder Bed Fusion (LPBF) is a method of additive manufacturing (AM) for metals that is increasingly being utilized in the aerospace industry. In addition to its comparative low buy to fly ratio, LPBF enables the design and manufacture of components that were previously not possible with subtractive manufacturing techniques. Due to the high temperatures achieved in laser melting of the powder, an inert shielding gas is used to protect the melt pool from contamination by oxygen, hydrogen and other potential reactive elements in the build chamber. The inert gas (commonly Argon) is introduced with nozzles to achieve a cross-flow over the build surface that whisks away byproducts generated by the laser-powder interaction. However, spatial variations or other undesirable characteristics (e.g., turbulence, dead zones, etc.) in the gas flow distribution could cause defects in the metal, spatial variability in quality and microstructure, as well as mechanical property variability. That “intra-build” variability can limit the application of LPBF and/or the application of parts produced by this technology for stress-critical applications. This investigation analyzed the inert gas flow in a full-scale model of the EOS M290 build chamber utilizing high-fidelity Particle Image Velocimetry (PIV) in planar mode using air as the gas medium. Results showed that the gas flow within the build chamber of the EOS M290 is not uniform across the build plate. Based on planar mode views parallel to the build plate, there is a reduction in the velocity from the left to the right side of approximately 47%. In addition, there is a gradient in the flow from the gas inlet to the exit baffle of approximately 66%. Although the highest cross-flow velocity is near the left side of the build plate, there is a dead zone in the back left corner with no tangential cross-flow. A recirculation zone was observed immediately downstream of the inlet vent, that spanned the build plate width, indicating that the gas cross-flow did not envelop the entire build plate. A recirculation region was also identified in planes perpendicular to the build plate that develops approximately midway between the upper and lower inlet nozzles. Vortices in the flow within that region could entrain smoke and metal vapor condensate, which in turn could cause attenuation of the incident laser. The flow showed spatial variations in turbulent kinetic energy, which could serve as a useful parameter to assess particle entrainment. These results provide new understanding that supports the design of components by LPBF with improved quality, and that could aid in the development of new nozzle geometry designs for the EOS M290 to optimize the gas flow.