Achieving exceptional strength-ductility synergy in a 3D-printed CrCoNi medium-entropy alloy with machine-learning assistance

Optimization of processing parameters is of great importance in additive manufacturing (AM) of metal alloys. Conventional trial-and-error approach is time- and cost-consumable, and the optimized parameters are sometimes sub-optimal, leading to the strength and ductility is not desirable. In this work, we employed Gaussian Process Regression machine learning (ML) to quickly find the optimized parameters for AM of CrCoNi MEA with relative density higher than 99 %. Using the optimized parameters, the additively manufactured CrCoNi MEA exhibited a tensile strength of 743 MPa and ductility of 59.5 %. The combination of the strength and ductility is 44.21GPa%, exceeding that of other CrCoNi MEA fabricated by AM without ML assistance. Such exceptional strength-ductility synergy was attributed to the original microstructures featuring fine grains, high dislocation density, and the deformed microstructural defects of twins, nanoscale network of stacking faults, dislocations, and the interactions between these defects. The method in this work sheds new sights into optimization of AM processing parameters and producing metal alloys with great strength-ductility synergy.