Data-driven optimisation of laser powder bed fusion–processed biodegradable Fe35Mn

Laser powder bed fusion (LPBF) enables the fabrication of patient-specific, porous, and biodegradable metallic implant architectures. However, robust processing windows remain challenging to define for biodegradable Fe–Mn alloys due to the intricate dependencies of density, surface finish, tensile properties, and corrosion response on coupled processing parameters. Here, 200 LPBF parameter combinations for Fe35Mn were generated using Latin hypercube sampling over laser power, scan speed, and hatch spacing. Relative density, top and side areal roughness, and tensile properties were then evaluated. Hierarchical Gaussian process surrogate models and multi-objective optimisation were then used to identify optimised processing parameters that balanced density, surface quality, and tensile performance. The resulting process map revealed that volumetric energy density captured the broad transition from insufficient fusion to near-dense material, but it is not sufficient to identify an optimal condition once densification had saturated. Even within the near-dense regime, surface roughness and elongation also remained strongly parameter-sensitive. The optimised processing parameters proposed here can produce tensile strengths up to 490 MPa and elongations up to 45%. Moreover, higher build rates were feasible while retaining tensile strength, but they increased surface roughness and reduced ductility, indicating that the practical productivity limit is governed mainly by surface quality and elongation. Electrochemical testing showed a corrosion rate of approximately 0.20 mm/year across the optimised conditions. Overall, the framework provides a practical LPBF process window for Fe35Mn and supports further evaluation of optimised LPBF-processed Fe35Mn for biodegradable implant applications.