Hits:

Affiliation of Author(s):能源与动力学院

Journal:Joint Thermophys. Heat Transf. Conf.

Abstract:In order to improve the film cooling performance, shape optimization of a round-to-slot hole was carried out. The exit slot width, the height and the injection angle of film hole were chosen as the design variables, and the spatially-averaged film cooling effectiveness was considered as objective function, which was to be maximized. The CFD results calculated by the Reliable k−Ε turbulence models were tested by the experimental data. Radial basis function neural network (RBFNN) was applied for surrogate model, and the optimal design point was determined by a kind of genetic algorithms. The CFD calculation results of 25 design points selected by Latin Hypercube Sampling (LHS) were used as training samples and the CFD calculation results of 8 design points selected randomly were used as testing samples. At M(the blowing ratio) =0.5, the optimal value of the exit slot width, the height and the injection angle of film hole are respectively 1.1 mm, 3.0 mm and 20°, which leading to 4% increase of spatially-averaged film cooling effectiveness. At M=1.5, the optimal value of the exit slot width, the height and the injection angle of film hole are respectively 2.2 mm, 3.9 mm and 29.1°, which leading to 23% increase of spatially-averaged film cooling effectiveness. Radial basis function neural network coupling with genetic algorithm is an effective tool to improve the film cooling performance. © 2018 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

Translation or Not:no

Date of Publication:2018-01-01

Co-author:Ying, Huang,wangchunhua

Correspondence Author:zjz