The combined use of order tracking techniques for enhanced Fourier analysis of order components

Abstract

Order tracking is one of the most important vibration analysis techniques for diagnosing faults in rotating machinery. It can be performed in many different ways, each of these with distinct advantages and disadvantages. However, in the end the analyst will often use Fourier analysis to transform the data from a time series to frequency or order spectra. It is therefore surprising that the study of the Fourier analysis of order-tracked systems seems to have been largely ignored in the literature. This paper considers the frequently used Vold-Kalman filter-based order tracking and computed order tracking techniques. The main pros and cons of each technique for Fourier analysis are discussed and the sequential use of Vold-Kalman filtering and computed order tracking is proposed as a novel idea to enhance the results of Fourier analysis for determining the order components. The advantages of the combined use of these order tracking techniques are demonstrated numerically on an SDOF rotor simulation model. Finally, the approach is also demonstrated on experimental data from a real rotating machine.

Detecting Vegetation Variations and Main Drivers over the Agropastoral Ecotone of Northern China through the Ensemble Empirical Mode Decomposition Method

“…In addition to the traditional linear regression method, nonlinear fitting methods have been regarded as a new way to analyze vegetation dynamics, including Fourier analysis [18], wavelet-based methods [19], breaks for additive season and trend (BFAST) [20], and detecting breakpoints and estimating segments in trend (DBEST) [16]. Although these tools were developed to detect changes in the vegetation variation trend, Fourier analysis mainly focuses on annual fluctuation, and the dataset to be analyzed should be absolutely periodic or stationary [21]; wavelet analysis is sensitive to noise; and trend detection methods are limited by the linear and nonlinear assumptions of signals [22]. Meanwhile, BFAST analysis predigests nonlinear trend components into some trend sections [23], and the intrinsic trend cannot be expressed; and DBEST analyses are more sensitive to trend changes with short durations, such as trend changes induced by wildfires and recovery [24].…”

The application of order tracking for vibration analysis of a varying speed rotor with a propagating transverse crack

“…Wang and Heyns [15] recently proposed the combined use of VKF-OT and COT in sequence, where the enhanced Fourier analysis order components are presented. However, the order tracking application in this paper is different from the sequential use of the two methods considered in [15], but rather uses each technique on its own for different purposes. A graphic explanation of the above discussion is given in Figure 1.…”

A multi-order probabilistic approach for Instantaneous Angular Speed tracking debriefing of the CMMNO׳14 diagnosis contest

“…It has been for long an essential quantity in rotor dynamics and in NVH applications concerned with the testing of rotating machines in transient regimes such as runups and shutdowns. In this context, the IAS makes possible the synchronization of Fourier analysis on the shaft rotation -a practice known as order (spectral) analysis -or the tracking of the evolution of the Fourier coefficients as a function of speed -a practice known as order tracking [5][6][7]. Early successes of order tracking are exemplified by the Vold-Kalman filter [8][9][10][11][12][13][14], a model-based approach well suited to slow-speed variations and high SNRs.…”

“…These aspects are discussed in Refs. [31][32][33][34][35]5]. Technological aspects of the measurement of the IAS and associated errors are further investigated in Refs.…”