Grey-RBF-FA method to optimize surface integrity for inclined end milling Inconel 718

Performance and service life of aerospace component are significantly influenced by surface integrity, especially for cyclically loaded thin-walled structure component. This work focuses on the surface roughness and residual stress of machined surface, which are considered the most important indications of surface integrity, for inclined end milling Ni-based superalloy Inconel 718. The purpose is to minimize surface roughness and maximize compressive residual stress by optimizing the cutter geometry. Based on the grey relational analysis (GRA), an integrated multi-objective optimization approach with the radial basis function (RBF) neural network and the firefly algorithm (FA) is developed. The end-tooth rake angle, end-tooth relief angle, and helix angle are selected as design factors, while the residual stress and the 3D surface roughness are taken as performance characteristics. The orthogonal array L16 (4³) is employed to generate sample data set. Then, the GRA is introduced to search the most influential factor by the grey relational grade (GRG). Subsequently, the proposed grey-RBF-FA method is applied to the multi-objective optimization problem. After calculating the GRG, the RBF network is used to relate the GRG with cutter geometric parameters. The test data show the RBF model has a low prediction error of 11.11%. Finally, the FA is utilized to search the optimal parameter-level combination. Validation experiments show that, compared with the original GRA, the proposed method further greatly improves the surface roughness and residual stresses in both directions by 1.20 μm, 249.1 MPa, and 176.5 MPa, respectively. The developed approach is proved to be feasible and can be generalized for other multi-objective optimization in manufacturing industry.