NUMERICAL OPTIMIZATION OF A STATOR VANE SETTING IN MULTISTAGE AXIAL-FLOW COMPRESSORS

This paper presents a numerical methodology for optimizing a stator st agger setting in a multistage axial-flow compressor environment. The m ethod involves simultaneous resetting of several blade rows, which inf luences overall performance in a complex manner. The paper is presente d in four parts: modelling the effect of stagger setting on individual stage performance, overall performance including surge point predicti on, stagger setting optimization and numerical examples. The stage per formance model is a one-dimensional meanline method where correlations are used to introduce real flow effects. The method uses (experimenta l or predicted) stage characteristics at design (nominal) setting to g enerate characteristics at other settings. A stage-by-stage model is u sed to 'stack' the stages together with a dynamic surge prediction mod el. A direct search method incorporating a sequential weight increasin g factor technique (SWIFT) was then used to optimize stagger setting. The objective function in this optimization is penalized externally wi th an updated factor which helped to accelerate convergence. The metho dology has been incorporated into a FORTRAN program and validated usin g data from a seven-stage aircraft compressor with hypothetical variab le stagger vanes. Results have demonstrated that variable stagger is a powerful method to rematch stages which can be used to improve desire d overall performance. Parametric studies on the optimization algorith m have also been conducted where it showed numerical stability and fas t convergence.