Abstract
The vaned diffuser profile designed preliminarily has been optimized using artificial neural network and genetic algorithm under stage environment to improve the aerodynamic performance at the design operation condition. The aerodynamic performance of the stage composed of the vaned diffuser optimized and the original impeller at the design and off-design operation conditions is predicted using the numerical method. The results show that the static pressure recovery coefficient of the vaned diffuser and the isentropic efficiency of the stage was increased by 11.7% and 1.64%, respectively. The total pressure loss coefficient of the vaned diffuser was decreased by 21.12%, and the isentropic efficiency of the stage was 86.1% at the design operation condition. Moreover, the aerodynamic performances of the stage at the off-design operation conditions was improved significantly. In addition, the surge margin of the stage was increased, and the choke margin was decreased slightly.
| Original language | English |
|---|---|
| Pages (from-to) | 32-36 |
| Number of pages | 5 |
| Journal | Hsi-An Chiao Tung Ta Hsueh/Journal of Xi'an Jiaotong University |
| Volume | 43 |
| Issue number | 11 |
| State | Published - Nov 2009 |
Keywords
- Aerodynamic optimization
- Artificial neural network
- Centrifugal compressor
- Genetic algorithm
- Vaned diffuser