Multi-objective optimization of aeroengine rotor assembly based on tensor coordinate transformation and NSGA-II

Abstract

In the stacked assembly process of the multi-stage rotor of an aeroengine, the cumulative error transmission caused by the geometric error and centroid deviation of a single blisk have a serious impact on the assembly quality and operation safety of the aeroengine rotor. Therefore, it is necessary to optimize the stacked assembly of the multi-stage rotor to improve assembly quality. In this paper, the Pareto-optimal schemes combining multi-objective optimization algorithms for stacked assembly of multi-stage rotors are proposed. First, the error propagation model for stacked assembly of multi-stage blisks based on tensor coordinate transformation is deduced. Taking blisk assembly phase as the design variable, the mathematical descriptions of geometric error and unbalance of the multi-stage rotor assembly are then established. The stacked assembly problem of the multi-stage rotor is formulated mathematically as a multi-objective optimization problem. The minimum geometric error and the minimum unbalance of multi-stage rotor are considered as optimization objectives. The Pareto-optimal assembly schemes are then solved based on the elitist non-dominated sorting genetic algorithm (NSGA-II). The simulation analysis and experimental study for the Pareto-optimal assembly schemes of multi-stage rotor are performed. Results show that compared with random stacked assembly, the geometric error is reduced by 40.6%, and the initial unbalance is reduced by 35.7%. The Pareto-optimal assembly schemes proposed in this paper can effectively improve stacked assembly quality of multi-stage aeroengine rotors.