A new method to predict the transport of non-uniform total pressure distributions through an axial flow compressor is presented. The method relies on frequency-domain transformations of total pressure distortion patterns, and the use of digital filter techniques to capture the effect of a blade row on the total pressure distortion. Compressor characteristics, described by a frequency response function, are obtained from experimental data and are related to fundamental blade row flow phenomena. When excited by a total pressure distribution with unknown response, the blade row's frequency response function can serve to predict the downstream total pressure profile.

This work covers the application of frequency response functions for turbomachinery blade rows, and includes the development of a process that improves the capability of linear response functions to model nonlinear flow behavior. The process is empirically based and relies on several original correlation parameters. Developed frequency domain modeling methods are applied to a rotor of modern design with promising results.