Application of bispectral analysis of vibroacoustic signals for diagnosis blades cracks in aircraft engines

Abstract

Background . The work is devoted to vibroacoustical diagnosis of the crack-like damages of the gas-turbine engines (GTE) blades at the steady-state modes of GTE. Objective . The effectiveness of the usage bispectrum analysis is investigated for signal processing at the vibroacoustical diagnostic of the blades cracks of aviation gas-turbine engines. Methods . Bispectral analysis is used for diagnostic information processing. Bispectrum can be found using Fourier transform third-order autocorrelation functions: \[S_{3x}(f_{1},f_{2})=\sum_{k,l=-\infty }^{\infty}C_{3x}(k,l)\textrm{exp}(-j2\pi f_{1}k)\textrm{exp}(-j2\pi f_{2}l).\] Results . Simulation bispectrum analysis vibroacoustical signals, radiated by models blades of working wheels without crack and with crack at the stationary vibrating excitations are carried out and investigated. It demonstrates that appearence and evolution of the fatigue crack in a blade increases intensity of bispectral modules. Conclusions . Application of a modern signal processing methods allows to detect fault features, which are sensitive to small crack-like damages. The receive results can be used to create a vibroacoustical monitoring system for aircraft engine rotor components.