面向装配精度的航空发动机转子零件选配优化

Aiming at the selection problem of aero-engine rotor parts, an approach to rotor part optimization for the stop-bolt connection structure was proposed by considering assembly requirements and individual characteristics of physical parts. Based on the measured size, shape and position error data of physical parts, a single rotor part error model was built for every part. According to the assembly relationship among parts, an assembly error propagation model of multi-stage rotors was established. By considering coaxiality and perpendicularity, a part matching optimization model was proposed. The fast elitist Non-dominated Sorting Genetic Algorithm (NSGA-Ⅱ) algorithm was used to optimize the part combination and circumferential assembly phase of each part. The exam showed that the coaxiality and perpendicularity errors were dramatically decreased compared to random part selection. By guiding part matching and assembly operation, the optimized result could improve success rate of single assembly try and consistency of assembly quality, and reduce trail-and-error cost.