Optimization of endwall profile for tail nozzle based on genetic algorithm

Abstract

In this paper, the endwall profile of the tail nozzle of a single stage transonic air turbine with a large expansion ratio is optimized to improve its aerodynamic performance under design conditions. The endwall profile optimization is carried out by using the design 3D module of NUMECA. The optimization selects seven parameters, including two control points of the shroud, four control points of the hub and the entrance angle. The pressure ratio and isentropic efficiency of the nozzle are took as the optimization objective. Artificial neural network simulation and genetic algorithm are combined to carry out the global optimization. Comparing the aerodynamic performance and the flow field analysis of the original and optimized nozzles, the results showed that the total pressure recovery coefficient of the nozzle is increased by 1.61%, the flow loss is reduced, and the aerodynamic performance of the nozzle is improved.