Digital twin construction and optimization design of disk-drum bolted structure for aero-engines

Abstract

In order to reduce the production, testing, and redesign cycles in the development process of aero-engine rotor systems, the digital twin (DT) for the disk-drum bolted structure (DDBS) of aero-engine is established. The DT encompasses geometric model, physical model, behavioral model, and rule model. The thin-layer element method is employed to characterize contact stiffness and the behavioral model that accurately reflects the physical entity is established. An adaptive genetic algorithm based on population diversity is applied to develop the rule model, which endows the DT with optimization capabilities. Finally, multi-objective structural optimization is conducted using the Analytic Hierarchy Process and DT, with the objectives of minimizing equivalent stress, reducing structural mass, and maximizing connection stiffness. Results indicate that the weights for connection stiffness, equivalent stress, and structural mass are 0.4737, 0.3233, and 0.2030, respectively. The optimized DDBS performance is significantly improved, with connection stiffness increased by 74.21%, and maximum equivalent stress and structural weight decreased by 15.67% and 7.25%, respectively. This research establishes a theoretical foundation for enhancing disk-drum bolted structure performance and provides novel insights and methodologies for optimizing other components and overall structures of aero-engines.