Laser polishing of additive manufactured stainless-steel parts by line focused beam: A response surface method for improving surface finish

Abstract

Laser polishing is a non-contact method in which laser energy is supplied to the surface selectively, causing localized melting and reflow of the material. A study was conducted on laser powder bed fusion (L-PBF) fabricated 15-5 precipitation hardening (PH) stainless steel specimens to examine the effect of the leading laser polishing process factors on the surface finish. A line-focused beam was employed to polish a large area quickly and effectively. A response surface methodology-based design of experiment was used, the regression equations were obtained, and various surface texture indicators were investigated to understand better the process mechanisms, further supported by a comprehensive microstructure analysis and microhardness test. The results showed a 60 % improvement in surface finish from an initial overall sample area Sa (leveled) of 14.1 μm to 5.57 μm. Gaussian and Robust Gaussian filter with a standard 0.8 mm cutoff was used to extract waviness and roughness. There was a dominance of form error and waviness at higher laser energy densities; thus, it can be shown that shallow surface melting (SSM) predominates at low laser energy densities, while surface over melt (SOM) dominates at higher laser energy densities. The minimum Ra value was obtained as 1.26 μm in the direction of the laser scan direction and 0.68 μm across the laser scan direction. The microhardness of the polished sample increased slightly compared to the base material. Microstructural examination showed no noticeable phase changes throughout the low-energy density laser polishing. At high energy density, the electron backscatter diffraction (EBSD) phase map revealed a gradient microstructure with the austenite content high at the bottom part of the solidified melt pool. Multi-objective desirability function-based optimization was done for minimum Sa (leveled) surface with minimum form error Sa (form). The confirmatory experiments validated the results within 13 % to 20 % error variation.