Applicability of Principle Components Analysis (PCA) to evaluate the dynamic characteristics of cavitation from captured images

Abstract

Cavitation in the flapper-nozzle pilot valve has been found as the mechanism of performance deterioration of the electro-hydraulic servo-valve. In this paper the transient cavitation patterns of the flapper-nozzle structure have been studied experimentally based on the Principle Components Analysis (PCA) method. Firstly the basic principle of PCA is introduced. Then a set of test images is used to invalidate the PCA method showing that PCA is able to predict the periodic variations among a series of images by investigating the coefficients of the principle components. Finally the cavitating flow field at different inlet pressures is captured using a high speed video camera. And the derived images have been analyzed to calculate the frequency of the coefficient variation of the first principle component. The results are also compared with the conclusions of the existing literature.