Optimising the Laser Powder Bed Fusion Process for the Manufacture of a Nickel Superalloy

Laser Powder bed Fusion as a form of additive manufacturing is a potentially desirable method of manufacture for aircraft parts that use nickel superalloys. One of the barriers to greater adoption is the uncertainty in mechanical properties build quality and microstructure that can arise from the LPBF process. This project explores how this uncertainty can be reduced or eliminated. Different methods achieve this reduction in uncertainty. A method for characterising both the laser and scanner system is outlined and tested. The different melting modes that can occur between the laser and the material are investigated. A closed-loop control system that included a machine learning algorithm was successfully implemented. This control had a statistically significant effect on reducing the variation in thermal emissivities. Constraints found from experimental observations and fundamental physics were imposed to reduce the creation of defects. The mechanical performance of Haynes 282 was benchmarked against other manufacturing methods and other alloys to find that test samples built using LPBF performed as well as wrought or cast samples in elevated stress rupture testing. In order to improve mechanical properties such as high temperature stress rupture resistance, the susceptibility of Haynes 282 to different cracking mechanisms was investigated. This susceptibility was tested through models and analysis of the microstructure and cracks themselves. The susceptibility was tested at different beam velocities, and it was found that the thermal front velocity was found to be critical in influencing the likelihood of crack formation. When the results of these experiments are combined, it suggests parameter ranges that suit both the alloy and the machine, whilst the closed-loop control can stay within these ranges even when the geometry or process conditions vary. This project was funded by GKN Aerospace in conjunction with the Engineering and Physical Sciences Research Council (EPSRC).