Dual-sided co-machining scheme optimization for thin-wall components with complex geometry

Addressing the challenge of deformation in thin-wall components due to insufficient rigidity, particularly during the machining of complex geometries, this paper draws on the digital twin framework to establish an optimization framework for dual-sided co-machining scheme. This method not only controls the milling forces on both sides to reduce deformation of the workpiece but also doubles the machining efficiency. In this paper, an overall framework with autonomous decision-making and evolutionary capabilities is introduced. Then, the practical effectiveness of the method is demonstrated through the complex task of machining propeller blades. Firstly, cutter path planning strategy and data processing method are devised. Next, in light of specific machining characteristics, implementation strategies are developed: an trial cutter path planning method based on the D-optimal design is devised using approximate periodicity to rapidly accumulate high-quality data; furthermore, a prediction function for deformation based on the neural network is developed through path relevance design. Once the prediction function is obtained, multi-objective optimization is performed to yield the most efficient machining scheme that meets the requirements for deformation. Finally, machining experiments confirm the effectiveness of this method in reducing deformation and enhancing machining efficiency.