Stochastic axial compressor variable geometry schedule optimisation

Date published

2011-07-01T00:00:00Z

Free to read from

Supervisor/s

Journal Title

Journal ISSN

Volume Title

Publisher

Elsevier Science B.V., Amsterdam.

Department

Type

Article

ISSN

1270-9638

Format

Citation

Abstract

The design of axial compressors is dictated by the maximisation of flow efficiency at on design conditions whereas at part speed the requirement for operation stability prevails. Among other stability aids, compressor variable geometry is employed to rise the surge line for the provision of an adequate surge margin. The schedule of the variable vanes is in turn typically obtained from expensive and time consuming rig tests that go through a vast combination of possible settings. The present paper explores the suitability of stochastic approaches to derive the most flow efficient schedule of an axial compressor for a minimum variable user defined value of the surge margin. A genetic algorithm has been purposely developed and its satisfactory performance validated against four representative benchmark functions. The work carries on with the necessary thorough investigation of the impact of the different genetic operators employed on the ability of the algorithm to find the global extremities in an effective and efficient manner. This deems fundamental to guarantee that the algorithm is not trapped in local extremities. The algorithm is then coupled with a compressor performance prediction tool that evaluates each individual's performance through a user defined fitness function. The most flow efficient schedule that conforms to a prescribed surge margin can be obtained thereby fast and inexpensively. Results are produced for a modern eight stage high bypass ratio compressor and compared with experimental data available to the research. The study concludes with the analysis of the existent relationship between surge margin and flow efficiency for the particular compressor under scrutiny. The study concludes with the analysis of the existent relationship between surge margin and flow efficiency for the particular compressor under scrutiny.

Description

Software Description

Software Language

Github

Keywords

Axial compressor, Compressor performance, Genetic algorithm, Optimisation, Variable geometry

DOI

Rights

Relationships

Relationships

Supplements

Funder/s