A study on the extraction of characteristics of compound faults of rolling bearings based on ITD-AF-CAF

Zhongyan An,Baodong Qiao,Mingyue Yu,Xiangdong Ge

doi:10.21595/jve.2021.21884

Abstract

In view of the cyclostationary characteristics of vibration signals from aero-engine, the combination of cyclic autocorrelation function and intrinsic timescale decomposition (ITD) has been proposed. According to the proposed method, vibration signals are decomposed by ITD algorithm to obtain the autocorrelation function of proper rotation components (PRC), based on which characteristic extraction and identification of compound faults of rolling bearings is made possible. To validate the effectiveness of method, an analysis has been given to the vibration signals of rolling bearings collected by sensors of different positions in different compound fault modes. As shown by results, the method combining ITD and cyclostationary theory can precisely and effectively extract the characteristic frequency relative to the type of faults and identify the compound faults.

Highlights

Rolling bearing is an important support part for rotation of aero-engine and extremely vulnerable working in high-speed and high-pressure environment
As found from the analysis of Fig. 4(c1), (c2), (c3), (c4) and Table 2, there is obvious characteristic frequency (58.4 Hz) of rolling elements and its frequency multiplication (701.6 Hz, 995.5 Hz, 1437 Hz) which can be found from the frequency spectrum of autocorrelation function of each rotation component after intrinsic timescale decomposition (ITD) decomposition, but without obvious characteristic frequency (148.8 Hz) of inner ring and its frequency multiplication component, which means that the method based on ITD-AF can partly extract the characteristic frequency of compound faults, but not comprehensively
Analyzing Fig. 6(a7), (b7) and Table 4, in case of compound faults of outer ring and a rolling element, the following conclusions can be drawn: No matter sensors are installed in horizontal direction or vertical, the proposed method can effectively extract the characteristic frequency 79 Hz of outer ring fault and triple characteristic frequency of rolling elements (157/3 = 52.3Hz) which matches with the type of faults

Summary

Introduction

Rolling bearing is an important support part for rotation of aero-engine and extremely vulnerable working in high-speed and high-pressure environment. Et al, proposed a quick algorithm based on intrinsic timescale decomposition [15] and the ITD algorithm was soon applied to the fault diagnosis of rotary machines, such as bearings and gears [16,17,18]. Victor Girondin brought forward the fault detection method for bearings of helicopter based on frequency adjustment and cyclostationary analysis [20]. A STUDY ON THE EXTRACTION OF CHARACTERISTICS OF COMPOUND FAULTS OF ROLLING BEARINGS BASED ON ITD-AF-CAF. The paper proposes the combination of ITD algorithm, correlation analysis and cyclostationary theory to extract characteristics of compound faults of rolling bearings and identify the type of faults

ITD-AF-CAF

Rolling bearing compound faults experiment

ITD-AF: common method

ITD-AF-CAF: a new method

Influencing factors

Conclusions