11th European Conference on Turbomachinery Fluid dynamics & Thermodynamics
To resolve secondary flows and wakes in turbomachinery measurements, reduction methods capable of isolating the deterministic and stochastic components of a given signal must be implemented. In the present study, two approaches are applied: a trigger-based phase-locked ensemble averaging technique and a newly proposed method based on the deconvolution of the signal in the frequency domain. While the former is used in numerous publications on the measurement of secondary flows, the latter has to these authors knowledge not yet been used in this context. Both data reduction techniques are outlined theoretically and applied to unsteady pressure measurements in a multi-stage axial turbine. Results obtained from the proposed deconvolution method are in good qualitative agreement with ensemble-averaged data. In some cases, a quantitative discrepancy is shown, most likely caused by numerical errors. Nevertheless, within certain boundaries the method is capable of isolating the deterministic signal including non-rotationally harmonic components without the need for a trigger signal.