The multi‐objective optimization of an axial cyclone separator in the gas turbine. (15th October 2021)

• Record Type: Journal Article
• Authors: Xing, Xiaolong; Pu, Wenhao; Zhang, Qi; Yang, Yu; Han, Dong
• Is Part Of: International journal of energy research. Volume 46:Number 3(2022)
• Page Start: 3428
• Page End: 3442
• Publication Date: 2021-10-15
• Subjects: axial cyclone separator -- blade outlet angles -- multi‐objective optimization -- nondominated sorting genetic algorithm II -- two‐fluid model; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Power resources -- Research -- Periodicals; 621.042
• DOI: 10.1002/er.7391
• Languages: English
• ISSNs: 0363-907X

Abstract

Summary: Herein the multi‐objective optimization of an axial cyclone separator is performed to enhance the overall performance with different velocities. And the separation efficiency and pressure drop are expected as the objective functions. In this paper, the difference value between internal and external blade outlet angles is regarded as one of the parameters to optimize. The regression expressions, which are obtained by Box‐Behnken Design and the two‐fluid model, are used in Nondominated Sorting Genetic Algorithm II to maximize separation efficiency and minimize pressure drop. It is found that the separation efficiency has a peak value with the rise of inlet velocity, and the response degrees of the internal angle, difference of the two angles, and blade number is disparate at different velocities. The pressure drops of the optimal axial cyclone tubes are smaller than the initial ones, but the separation efficiency is almost the same. Meanwhile, the accuracies of the optimization and numerical results are verified by the experiments. For the samples in 3, 5, and 7 m/s, pressure drop can be reduced by 24.1%, 22.7%, and 34.3% respectively. The simulation results indicate that the radial pressure difference and swirl number have been reduced. Besides, the optimal designs have more stable and even near wall particle distribution after multi‐objective optimization.