关键词: 伴随方法, 高速列车, 气动力, 外形优化, 网格变形

Abstract: In order to improve the aerodynamic performance of high-speed trains running in open air, an aerodynamic optimization design of a high-speed train head is carried out based on the adjoint method and the mesh deformation technique combining radial basic functions (RBF). The mesh deformation technique combining RBF is introduced to avoid repeated generation of the mesh during the optimization of the train head. Thus, the optimization efficiency has been improved. The sensitivity between the objective functions and the head shape is calculated through the adjoint method. This method don't need to define any design variables of the head shape, therefore, the effect of the artificial defined design variables on the optimization results can be avoided. An optimization process of the high-speed train head is proposed, combining the mesh deformation technique, the adjoint method and the computational fluid dynamic (CFD) method. The aerodynamic drag force of the whole train and the aerodynamic lift force of the tail car are set as the optimization objectives, and then the multi-objective aerodynamic optimization design of the high-speed train head is performed. The results show that the adjoint method can be efficiently applied to the optimization of the high-speed train head. After optimization under the constrains, the aerodynamic drag force of the whole train has been reduced by 2.83%, and the aerodynamic lift force of the tail car has been reduced by 25.86%. The reduction of the aerodynamic drag force is mainly located at the streamlined part of the head car and the tail car, and the aerodynamic drag forces of the middle car and the train body of the head car and the tail car basically keep unchanged. The reduction of the aerodynamic lift force of the tail car is mainly located at the streamlined part, and the absolute value in the down lift force of the train body of the tail car decreases slightly.

Key words: adjoint method, aerodynamic force, high-speed trains, mesh deformation, shape optimization