J. Wolff (Advisor)
Master of Science in Engineering (MSEgr)
A set of inlet guide vane (IGV) unsteady surface pressure measurements of a transonic compressor is presented. Using a flexible pressure sensor array, unsteady IGV suction-surface and pressure-surface pressures are acquired for six spanwise by five chordwise locations for various speed lines and throttle settings. Measurements from this sensor array are used to investigate unsteady vane/blade interaction aeromechanical forcing functions in a modern, highly loaded compressor stage. A significant effect is shown on the unsteady forced response of the IGV with changes in compressor operating point and IGV/rotor axial spacing for various span and chord locations. In particular, variations in the compressor operating point (i.e., mass flow rate and pressure ratio) cause change in both the magnitude and phase of the forced response, with the near-stall operating point producing the highest response. Changes in the axial spacing between the IGV and rotor rows from 12% to 26% of the IGV chord resulted in a 50% reduction in the magnitude of the forced response. A significant variation in the forced response with span is noted, especially at the 5% span location where the rotor relative flow is subsonic. In this region, changes in the operating point and axial spacing had a negligible effect on the forced response of the IGV. An innovative data reduction/analysis method is presented to quantify and statistically analyze the degree of blade-to-blade variations in the measured aerodynamic forcing functions obtained by turbomachinery experimentation. This method is used to analyze experimental data of IGV surface unsteady pressure response due to the aerodynamic forcing function produced by the downstream transonic compressor rotor with (1) factory-whole blades and (2) trimmed (blended) blades resulting from the repair of crack damage on two of the rotor blades. Results from the variation metric and l2-norm analysis indicate that the scaled metric possesses large magnitude change versus blade index for the trimmed rotor compared to that of the untrimmed rotor, with the largest values occurring near the trimmed blades. Each method is nearly always able to detect the trimmed blades. Using the distance between cluster centroids from the K-means cluster analysis as a metric of variation within each rotor, cluster distances increased by as much as a factor of 4 for the trimmed rotor compared to the untrimmed rotor. Therefore, correctly identifying the trimmed rotor data as having a significantly higher amount of blade-to-blade variability. Finally using the cluster distance as a goodness parameter for variability, the non-trimmed data was investigated for trends with changes in compressor operating conditions. This analysis showed an increase in blade-to-blade variability with increases in the compressor flow rate. Therefore, this data reduction/analysis method has the potential to be utilized as an indicator of the compressor operating point for control methods.
Department or Program
Department of Mechanical and Materials Engineering
Year Degree Awarded
Copyright 2003, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.