Mechanical anisotropy is a critical issue in laser powder bed fusion (LPBF) fabricated metallic materials due to directional solidification and heterogeneous microstructure development. In this work, A Gaussian Process Regression (GPR) model was employed to establish the correlation between process parameters and relative density, enabling efficient identification of a stable processing window with high densification of CoCrFeMnNi high-entropy alloy (HEA) through laser powder bed fusion (LPBF) additive manufacturing (AM). By modulating laser power and scanning speed, three distinct solidification microstructures—herringbone, bimodal, and columnar—were obtained. The observed differences in microstructure evolution are primarily attributed to epitaxial grain growth and side-branching phenomena occurring within the melt pool center and in regions where melt pools overlap. These three different microstructures exhibited diverse mechanical properties as well as various anisotropic tensile behavior along and perpendicular to the build direction. The optimal ultimate tensile strength of ∼628 MPa was revealed in the herringbone microstructure with a weak anisotropy in tensile property, while the bimodal microstructure exhibited a strong tensile anisotropy showing much higher ductility (i.e., ∼30%) in the vertical condition compared to ∼12% in the horizontal condition. The underlying mechanisms were analyzed with respect to effective grain size, heterogeneous microstructure, and the twinning-induced plasticity (TWIP) effect. Physics-based computational modeling and Gaussian process regression model-based machine learning techniques were employed to simulate the thermal profiles of the melt pool and assist the LPBF process optimization, respectively. This study demonstrates that controlled adjustment of LPBF processing parameters, guided by machine learning–assisted processing window mapping, provides an effective strategy for tailoring microstructure and regulating anisotropic mechanical behaviour in additively manufactured CoCrFeMnNi high-entropy alloys (HEAs). This approach provides insights into the design and fabrication of site-specific solidification microstructures with desired mechanical anisotropy by manipulating side-branching and epitaxial grain growth through varying laser powder bed fusion (LPBF) processing parameters.
Kong et al. (Wed,) studied this question.