Autocorrelation Ensemble Average of Larger Amplitude Impact Transients for the Fault Diagnosis of Rolling Element Bearings

Lei Hu, Yuandong Xu, Fengshou Gu, Jing He, Niaoqing Hu, Andrew Ball

Research output: Contribution to journalArticle

Abstract

Rolling element bearings are one of the critical elements in rotating machinery of energy engineering systems. A defective roller of bearing moves in and out of the load zone during each revolution of the cage. Larger amplitude impact transients (LAITs) are produced when the defective roller passes the load zone centre and the defective area strikes the inner or outer races. A series of LAIT segments with higher signal to noise ratio are separated from a continuous vibration signal according to the bearing geometry and kinematics. In order to eliminate the phase errors between different LAIT segments that can arise from rotational speed fluctuations and roller slippages, unbiased autocorrelation is introduced to align the phases of LAIT segments. The unbiased autocorrelation signals make the ensemble averaging more accurate, and hence, archive enhanced diagnostic signatures, which are denoted as LAIT-AEAs for brevity. The diagnostic method based on LAIT separation and autocorrelation ensemble average (AEA) is evaluated with the datasets captured from real bearings of two different experiment benches. The validation results of the LAIT-AEAs are compared with the squared envelope spectrums (SESs) yielded based on two state-of-the-art techniques of Fast Kurtogram and Autogram.

Original languageEnglish
Article number4740
Number of pages19
JournalEnergies
Volume12
Issue number24
DOIs
Publication statusPublished - 12 Dec 2019

Fingerprint

Bearings (structural)
Fault Diagnosis
Autocorrelation
Failure analysis
Ensemble
Diagnostics
Rotating Machinery
Phase Error
Rotating machinery
Vibration Signal
Cage
Systems Engineering
Systems engineering
Envelope
Averaging
Kinematics
Signal to noise ratio
Signature
Eliminate
Fluctuations

Cite this

@article{a5e242f6435d44419e05fc90b0bffc45,
title = "Autocorrelation Ensemble Average of Larger Amplitude Impact Transients for the Fault Diagnosis of Rolling Element Bearings",
abstract = "Rolling element bearings are one of the critical elements in rotating machinery of energy engineering systems. A defective roller of bearing moves in and out of the load zone during each revolution of the cage. Larger amplitude impact transients (LAITs) are produced when the defective roller passes the load zone centre and the defective area strikes the inner or outer races. A series of LAIT segments with higher signal to noise ratio are separated from a continuous vibration signal according to the bearing geometry and kinematics. In order to eliminate the phase errors between different LAIT segments that can arise from rotational speed fluctuations and roller slippages, unbiased autocorrelation is introduced to align the phases of LAIT segments. The unbiased autocorrelation signals make the ensemble averaging more accurate, and hence, archive enhanced diagnostic signatures, which are denoted as LAIT-AEAs for brevity. The diagnostic method based on LAIT separation and autocorrelation ensemble average (AEA) is evaluated with the datasets captured from real bearings of two different experiment benches. The validation results of the LAIT-AEAs are compared with the squared envelope spectrums (SESs) yielded based on two state-of-the-art techniques of Fast Kurtogram and Autogram.",
keywords = "Autocorrelation ensemble average, Fault diagnosis, Larger amplitude impact transient, Noise reduction, Roller slippage, Rolling element bearings, Rotational speed fluctuation",
author = "Lei Hu and Yuandong Xu and Fengshou Gu and Jing He and Niaoqing Hu and Andrew Ball",
year = "2019",
month = "12",
day = "12",
doi = "10.3390/en12244740",
language = "English",
volume = "12",
journal = "Energies",
issn = "1996-1073",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "24",

}

Autocorrelation Ensemble Average of Larger Amplitude Impact Transients for the Fault Diagnosis of Rolling Element Bearings. / Hu, Lei; Xu, Yuandong; Gu, Fengshou; He, Jing; Hu, Niaoqing; Ball, Andrew.

In: Energies, Vol. 12, No. 24, 4740, 12.12.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Autocorrelation Ensemble Average of Larger Amplitude Impact Transients for the Fault Diagnosis of Rolling Element Bearings

AU - Hu, Lei

AU - Xu, Yuandong

AU - Gu, Fengshou

AU - He, Jing

AU - Hu, Niaoqing

AU - Ball, Andrew

PY - 2019/12/12

Y1 - 2019/12/12

N2 - Rolling element bearings are one of the critical elements in rotating machinery of energy engineering systems. A defective roller of bearing moves in and out of the load zone during each revolution of the cage. Larger amplitude impact transients (LAITs) are produced when the defective roller passes the load zone centre and the defective area strikes the inner or outer races. A series of LAIT segments with higher signal to noise ratio are separated from a continuous vibration signal according to the bearing geometry and kinematics. In order to eliminate the phase errors between different LAIT segments that can arise from rotational speed fluctuations and roller slippages, unbiased autocorrelation is introduced to align the phases of LAIT segments. The unbiased autocorrelation signals make the ensemble averaging more accurate, and hence, archive enhanced diagnostic signatures, which are denoted as LAIT-AEAs for brevity. The diagnostic method based on LAIT separation and autocorrelation ensemble average (AEA) is evaluated with the datasets captured from real bearings of two different experiment benches. The validation results of the LAIT-AEAs are compared with the squared envelope spectrums (SESs) yielded based on two state-of-the-art techniques of Fast Kurtogram and Autogram.

AB - Rolling element bearings are one of the critical elements in rotating machinery of energy engineering systems. A defective roller of bearing moves in and out of the load zone during each revolution of the cage. Larger amplitude impact transients (LAITs) are produced when the defective roller passes the load zone centre and the defective area strikes the inner or outer races. A series of LAIT segments with higher signal to noise ratio are separated from a continuous vibration signal according to the bearing geometry and kinematics. In order to eliminate the phase errors between different LAIT segments that can arise from rotational speed fluctuations and roller slippages, unbiased autocorrelation is introduced to align the phases of LAIT segments. The unbiased autocorrelation signals make the ensemble averaging more accurate, and hence, archive enhanced diagnostic signatures, which are denoted as LAIT-AEAs for brevity. The diagnostic method based on LAIT separation and autocorrelation ensemble average (AEA) is evaluated with the datasets captured from real bearings of two different experiment benches. The validation results of the LAIT-AEAs are compared with the squared envelope spectrums (SESs) yielded based on two state-of-the-art techniques of Fast Kurtogram and Autogram.

KW - Autocorrelation ensemble average

KW - Fault diagnosis

KW - Larger amplitude impact transient

KW - Noise reduction

KW - Roller slippage

KW - Rolling element bearings

KW - Rotational speed fluctuation

UR - http://www.scopus.com/inward/record.url?scp=85077012720&partnerID=8YFLogxK

U2 - 10.3390/en12244740

DO - 10.3390/en12244740

M3 - Article

AN - SCOPUS:85077012720

VL - 12

JO - Energies

JF - Energies

SN - 1996-1073

IS - 24

M1 - 4740

ER -