Closed-loop Control of Melt Pool Temperature in Directed Energy Deposition Additive Manufacturing Using a Co-axial Two-wavelength Pyrometer

This work concerns the laser directed energy deposition (DED) additive manufacturing of stainless steel parts. In this paper we demonstrate that controlling the meltpool temperature through real-time closed-loop modulation of the laser power results in parts with uniform microstructure and reduced porosity. The DED process has the potential to replace traditional subtractive and formative manufacturing processes in rapid net-shape production, repair, and gradient multi-material applications. A major barrier in practical deployment of the DED process is its tendency to create flaws, such as porosity and non-uniformity (heterogeneity) of microstructure, which leads to large variation in functional properties. Therefore, process monitoring and closed-loop control in DED are urgently needed for ensuring repeatable part quality. In this work, we deposited six trapezoid-shaped stainless steel parts (SS 316L). Three of the six parts were produced with fixed a priori optimized process parameters, i.e., open-loop processing. Three additional parts were deposited with the same processing parameters as the open-loop parts, but the laser power was automatically and continually modulated via closed-loop control such that the meltpool temperature was maintained at 1650 °C. The meltpool temperature was measured in real-time using a two-wavelength pyrometer instrumented coaxial to the nozzle in the laser path. Subsequent to deposition, all six parts were characterized using X-ray computed tomography (porosity assessment), scanning electron microscopy (microstructure), and microhardness testing. Parts deposited under closed-loop control, termed Intelligent Processing depicted consistent dendritic microstructure. By contrast, parts built under fixed parameter, open-loop conditions or Standard Processing, showed microstructure heterogeneity with presence of both dendritic and planar grains, which leads to large variation in microhardness. Further, parts built under Intelligent Processing conditions have reduced deviation in percentage volume porosity ranging from 0.036% to 0.043%, while volume porosity for parts built under Standard Processing conditions ranged from 0.032% to 0.068%.