Sensitive data-driven tooth surface collaborative grinding model for aerospace spiral bevel gears

Abstract

To distinguish with the conventional grinding requiring geometric accuracy and loaded contact pattern (LCP), an innovative data-driven tooth surface collaborative grinding model is established by fine-modifying sensitive hypoid generator parameters. Especially, in sensitivity analysis-based adaptive prediction, robust control, an innovative adaptive material removal optimization considering loaded contact deformation of both normal and tangential directions is developed. It shows local constraint for LCP in tangential direction, and global constraint for geometric performance in normal direction. Moreover, in normal direction, the loaded contact deformation is also taken into account for an accurate prediction of tooth surface deformation. Then, the nonlinear adaptive prediction is performed by using universal motion concept (UMC) hypoid generator parameter modification. A sensitivity analysis strategy of hypoid generator parameters is used to perform adaptive robust control. Finally, Levenberg-Marquardt (L-M) method is adopted to solve the established control model for accurate optimal hypoid generator parameters with modification variations. The given test instance can verify the proposed method.